Machine learning for the prediction of bone metastasis in patients with newly diagnosed thyroid cancer

Machine Learning Algorithm for Predicting Distant Metastasis of T1 and T2 Gallbladder Cancer Based on SEER Database

(1) Background: This study seeks to employ a machine learning (ML) algorithm to forecast the risk of distant metastasis (DM) in patients with T1 and T2 gallbladder cancer (GBC); (2) Methods: Data of patients diagnosed with T1 and T2 GBC was obtained from SEER, encompassing the period from 2004 to 2015, were utilized to apply seven ML algorithms. These algorithms were appraised by the area under the receiver operating characteristic curve (AUC) and other metrics; (3) Results: This study involved 4371 patients in total. Out of these patients, 764 (17.4%) cases progressed to develop DM. Utilizing a logistic regression (LR) model to identify independent risk factors for DM of gallbladder cancer (GBC). A nomogram has been developed to forecast DM in early T-stage gallbladder cancer patients. Through the evaluation of different models using relevant indicators, it was discovered that Random Forest (RF) exhibited the most outstanding predictive performance; (4) Conclusions: RF has demonstrated high accuracy in predicting DM in gallbladder cancer patients, assisting clinical physicians in enhancing the accuracy of diagnosis. This can be particularly valuable for improving patient outcomes and optimizing treatment strategies. We employ the RF algorithm to construct the corresponding web calculator.

Artificial Intelligence in Detection, Management, and Prognosis of Bone Metastasis: A Systematic Review

BACKGROUND

Metastasis commonly occur in the bone tissue. Artificial intelligence (AI) has become increasingly prevalent in the medical sector as support in decision-making, diagnosis, and treatment processes. The objective of this systematic review was to assess the reliability of AI systems in clinical, radiological, and pathological aspects of bone metastases.

METHODS

We included studies that evaluated the use of AI applications in patients affected by bone metastases. Two reviewers performed a digital search on 31 December 2023 on PubMed, Scopus, and Cochrane library and extracted authors, AI method, interest area, main modalities used, and main objectives from the included studies.

RESULTS

We included 59 studies that analyzed the contribution of computational intelligence in diagnosing or forecasting outcomes in patients with bone metastasis. Six studies were specific for spine metastasis. The study involved nuclear medicine (44.1%), clinical research (28.8%), radiology (20.4%), or molecular biology (6.8%). When a primary tumor was reported, prostate cancer was the most common, followed by lung, breast, and kidney.

CONCLUSIONS

Appropriately trained AI models may be very useful in merging information to achieve an overall improved diagnostic accuracy and treatment for metastasis in the bone. Nevertheless, there are still concerns with the use of AI systems in medical settings. Ethical considerations and legal issues must be addressed to facilitate the safe and regulated adoption of AI technologies. The limitations of the study comprise a stronger emphasis on early detection rather than tumor management and prognosis as well as a high heterogeneity for type of tumor, AI technology and radiological techniques, pathology, or laboratory samples involved.

Comparison of Two Modern Survival Prediction Tools, SORG-MLA and METSSS, in Patients With Symptomatic Long-bone Metastases Who Underwent Local Treatment With Surgery Followed by Radiotherapy and With Radiotherapy Alone

BACKGROUND

Survival estimation for patients with symptomatic skeletal metastases ideally should be made before a type of local treatment has already been determined. Currently available survival prediction tools, however, were generated using data from patients treated either operatively or with local radiation alone, raising concerns about whether they would generalize well to all patients presenting for assessment. The Skeletal Oncology Research Group machine-learning algorithm (SORG-MLA), trained with institution-based data of surgically treated patients, and the Metastases location, Elderly, Tumor primary, Sex, Sickness/comorbidity, and Site of radiotherapy model (METSSS), trained with registry-based data of patients treated with radiotherapy alone, are two of the most recently developed survival prediction models, but they have not been tested on patients whose local treatment strategy is not yet decided.

QUESTIONS/PURPOSES

(1) Which of these two survival prediction models performed better in a mixed cohort made up both of patients who received local treatment with surgery followed by radiotherapy and who had radiation alone for symptomatic bone metastases? (2) Which model performed better among patients whose local treatment consisted of only palliative radiotherapy? (3) Are laboratory values used by SORG-MLA, which are not included in METSSS, independently associated with survival after controlling for predictions made by METSSS?

METHODS

Between 2010 and 2018, we provided local treatment for 2113 adult patients with skeletal metastases in the extremities at an urban tertiary referral academic medical center using one of two strategies: (1) surgery followed by postoperative radiotherapy or (2) palliative radiotherapy alone. Every patient's survivorship status was ascertained either by their medical records or the national death registry from the Taiwanese National Health Insurance Administration. After applying a priori designated exclusion criteria, 91% (1920) were analyzed here. Among them, 48% (920) of the patients were female, and the median (IQR) age was 62 years (53 to 70 years). Lung was the most common primary tumor site (41% [782]), and 59% (1128) of patients had other skeletal metastases in addition to the treated lesion(s). In general, the indications for surgery were the presence of a complete pathologic fracture or an impending pathologic fracture, defined as having a Mirels score of ≥ 9, in patients with an American Society of Anesthesiologists (ASA) classification of less than or equal to IV and who were considered fit for surgery. The indications for radiotherapy were relief of pain, local tumor control, prevention of skeletal-related events, and any combination of the above. In all, 84% (1610) of the patients received palliative radiotherapy alone as local treatment for the target lesion(s), and 16% (310) underwent surgery followed by postoperative radiotherapy. Neither METSSS nor SORG-MLA was used at the point of care to aid clinical decision-making during the treatment period. Survival was retrospectively estimated by these two models to test their potential for providing survival probabilities. We first compared SORG to METSSS in the entire population. Then, we repeated the comparison in patients who received local treatment with palliative radiation alone. We assessed model performance by area under the receiver operating characteristic curve (AUROC), calibration analysis, Brier score, and decision curve analysis (DCA). The AUROC measures discrimination, which is the ability to distinguish patients with the event of interest (such as death at a particular time point) from those without. AUROC typically ranges from 0.5 to 1.0, with 0.5 indicating random guessing and 1.0 a perfect prediction, and in general, an AUROC of ≥ 0.7 indicates adequate discrimination for clinical use. Calibration refers to the agreement between the predicted outcomes (in this case, survival probabilities) and the actual outcomes, with a perfect calibration curve having an intercept of 0 and a slope of 1. A positive intercept indicates that the actual survival is generally underestimated by the prediction model, and a negative intercept suggests the opposite (overestimation). When comparing models, an intercept closer to 0 typically indicates better calibration. Calibration can also be summarized as log(O:E), the logarithm scale of the ratio of observed (O) to expected (E) survivors. A log(O:E) > 0 signals an underestimation (the observed survival is greater than the predicted survival); and a log(O:E) < 0 indicates the opposite (the observed survival is lower than the predicted survival). A model with a log(O:E) closer to 0 is generally considered better calibrated. The Brier score is the mean squared difference between the model predictions and the observed outcomes, and it ranges from 0 (best prediction) to 1 (worst prediction). The Brier score captures both discrimination and calibration, and it is considered a measure of overall model performance. In Brier score analysis, the "null model" assigns a predicted probability equal to the prevalence of the outcome and represents a model that adds no new information. A prediction model should achieve a Brier score at least lower than the null-model Brier score to be considered as useful. The DCA was developed as a method to determine whether using a model to inform treatment decisions would do more good than harm. It plots the net benefit of making decisions based on the model's predictions across all possible risk thresholds (or cost-to-benefit ratios) in relation to the two default strategies of treating all or no patients. The care provider can decide on an acceptable risk threshold for the proposed treatment in an individual and assess the corresponding net benefit to determine whether consulting with the model is superior to adopting the default strategies. Finally, we examined whether laboratory data, which were not included in the METSSS model, would have been independently associated with survival after controlling for the METSSS model's predictions by using the multivariable logistic and Cox proportional hazards regression analyses.

RESULTS

Between the two models, only SORG-MLA achieved adequate discrimination (an AUROC of > 0.7) in the entire cohort (of patients treated operatively or with radiation alone) and in the subgroup of patients treated with palliative radiotherapy alone. SORG-MLA outperformed METSSS by a wide margin on discrimination, calibration, and Brier score analyses in not only the entire cohort but also the subgroup of patients whose local treatment consisted of radiotherapy alone. In both the entire cohort and the subgroup, DCA demonstrated that SORG-MLA provided more net benefit compared with the two default strategies (of treating all or no patients) and compared with METSSS when risk thresholds ranged from 0.2 to 0.9 at both 90 days and 1 year, indicating that using SORG-MLA as a decision-making aid was beneficial when a patient's individualized risk threshold for opting for treatment was 0.2 to 0.9. Higher albumin, lower alkaline phosphatase, lower calcium, higher hemoglobin, lower international normalized ratio, higher lymphocytes, lower neutrophils, lower neutrophil-to-lymphocyte ratio, lower platelet-to-lymphocyte ratio, higher sodium, and lower white blood cells were independently associated with better 1-year and overall survival after adjusting for the predictions made by METSSS.

CONCLUSION

Based on these discoveries, clinicians might choose to consult SORG-MLA instead of METSSS for survival estimation in patients with long-bone metastases presenting for evaluation of local treatment. Basing a treatment decision on the predictions of SORG-MLA could be beneficial when a patient's individualized risk threshold for opting to undergo a particular treatment strategy ranged from 0.2 to 0.9. Future studies might investigate relevant laboratory items when constructing or refining a survival estimation model because these data demonstrated prognostic value independent of the predictions of the METSSS model, and future studies might also seek to keep these models up to date using data from diverse, contemporary patients undergoing both modern operative and nonoperative treatments.

LEVEL OF EVIDENCE

Level III, diagnostic study.

Prognostic and predictive value of examined lymph node count in stage III colorectal cancer: a population based study

BACKGROUND

The role of tumor-draining lymph nodes in the progression of malignant tumors, including stage III colorectal cancer (CRC), is critical. However, the prognostic and predictive value of the number of examined lymph nodes (ELNs) are not fully understood.

METHODS

This population-based study retrospectively analyzed data from 106,843 patients with stage III CRC who underwent surgical treatment and registered in three databases from 2004 to 2021. The Surveillance, Epidemiology, and End Results (SEER) cohort was divided using into training and test cohorts at a ratio of 3:2. We employed restricted cubic spline (RCS) curves to explore nonlinear relationships between overall survival (OS) and ELNs counts and performed Cox regression to evaluate hazard ratios across different ELNs count subtypes. Additional validation cohorts were utilized from the First Affiliated Hospital, Sun Yat-sen University and The Cancer Genome Atlas (TCGA) under the same criteria. Outcomes measured included OS, cancer-specific survival (CSS), and progression-free survival (PFS). Molecular analyses involved differential gene expression using the "limma" package and immune profiling through CIBERSORT. Tissue microarray slides and multiplex immunofluorescence (MIF) were used to assess protein expression and immune cell infiltration.

RESULTS

Patients with higher ELNs counts (≥ 17) demonstrated significantly better long-term survival outcomes across all cohorts. Enhanced OS, CSS, and PFS were notably evident in the LN-ELN group compared to those with fewer ELNs. Cox regression models underscored the prognostic value of higher ELNs counts across different patient subgroups by age, sex, tumor differentiation, and TNM stages. Subtype analysis based on ELNs count revealed a marked survival benefit in patients treated with adjuvant chemotherapy in the medium and large ELNs counts (≥ 12), whereas those with fewer ELNs showed negligible benefits. RNA sequencing and MIF indicated elevated immune activation in the LN-ELN group, characterized by increased CD3+, CD4+, and CD8 + T cells within the tumor microenvironment.

CONCLUSIONS

The number of ELNs independently predicts survival and the immunological landscape at the tumor site in stage III CRC, underscoring its dual prognostic and predictive value.

Machine learning based on SEER database to predict distant metastasis of thyroid cancer

OBJECTIVE

Distant metastasis of thyroid cancer often indicates poor prognosis, and it is important to identify patients who have developed distant metastasis or are at high risk as early as possible. This paper aimed to predict distant metastasis of thyroid cancer through the construction of machine learning models to provide a reference for clinical diagnosis and treatment.

MATERIALS & METHODS

Data on demographic and clinicopathological characteristics of thyroid cancer patients between 2010 and 2015 were extracted from the National Institutes of Health (NIH) Surveillance, Epidemiology, and End Results (SEER) database. Our research used univariate and multivariate logistic models to screen independent risk factors, respectively. Decision Trees (DT), ElasticNet (ENET), Logistic Regression (LR), Extreme Gradient Boosting (XGBoost), Random Forest (RF), Multilayer Perceptron (MLP), Radial Basis Function Support Vector Machine (RBFSVM) and seven machine learning models were compared and evaluated by the following metrics: the area under receiver operating characteristic curve (AUC), calibration curve, decision curve analysis (DCA), sensitivity(also called recall), specificity, precision, accuracy and F1 score. Interpretable machine learning was used to identify possible correlation between variables and distant metastasis.

RESULTS

Independent risk factors for distant metastasis, including age, gender, race, marital status, histological type, capsular invasion, and number of lymph nodes metastases were screened by multifactorial regression analysis. Among the seven machine learning algorithms, RF was the best algorithm, with an AUC of 0.948, sensitivity of 0.919, accuracy of 0.845, and F1 score of 0.886 in the training set, and an AUC of 0.960, sensitivity of 0.929, accuracy of 0.906, and F1 score of 0.908 in the test set.

CONCLUSIONS

The machine learning model constructed in this study helps in the early diagnosis of distant thyroid metastases and helps physicians to make better decisions and medical interventions.

Interpretable machine learning-based clinical prediction model for predicting lymph node metastasis in patients with intrahepatic cholangiocarcinoma

OBJECTIVE

Prediction of lymph node metastasis (LNM) for intrahepatic cholangiocarcinoma (ICC) is critical for the treatment regimen and prognosis. We aim to develop and validate machine learning (ML)-based predictive models for LNM in patients with ICC.

METHODS

A total of 345 patients with clinicopathological characteristics confirmed ICC from Jan 2007 to Jan 2019 were enrolled. The predictors of LNM were identified by the least absolute shrinkage and selection operator (LASSO) and logistic analysis. The selected variables were used for developing prediction models for LNM by six ML algorithms, including Logistic regression (LR), Gradient boosting machine (GBM), Extreme gradient boosting (XGB), Random Forest (RF), Decision tree (DT), Multilayer perceptron (MLP). We applied 10-fold cross validation as internal validation and calculated the average of the areas under the receiver operating characteristic (ROC) curve to measure the performance of all models. A feature selection approach was applied to identify importance of predictors in each model. The heat map was used to investigate the correlation of features. Finally, we established a web calculator using the best-performing model.

RESULTS

In multivariate logistic regression analysis, factors including alcoholic liver disease (ALD), smoking, boundary, diameter, and white blood cell (WBC) were identified as independent predictors for LNM in patients with ICC. In internal validation, the average values of AUC of six models ranged from 0.820 to 0.908. The XGB model was identified as the best model, the average AUC was 0.908. Finally, we established a web calculator by XGB model, which was useful for clinicians to calculate the likelihood of LNM.

CONCLUSION

The proposed ML-based predicted models had a good performance to predict LNM of patients with ICC. XGB performed best. A web calculator based on the ML algorithm showed promise in assisting clinicians to predict LNM and developed individualized medical plans.

Development and validation of prediction models for papillary thyroid cancer structural recurrence using machine learning approaches

BACKGROUND

Although papillary thyroid cancer (PTC) patients are known to have an excellent prognosis, up to 30% of patients experience disease recurrence after initial treatment. Accurately predicting disease prognosis remains a challenge given that the predictive value of several predictors remains controversial. Thus, we investigated whether machine learning (ML) approaches based on comprehensive predictors can predict the risk of structural recurrence for PTC patients.

METHODS

A total of 2244 patients treated with thyroid surgery and radioiodine were included. Twenty-nine perioperative variables consisting of four dimensions (demographic characteristics and comorbidities, tumor-related variables, lymph node (LN)-related variables, and metabolic and inflammatory markers) were analyzed. We applied five ML algorithms-logistic regression (LR), support vector machine (SVM), extreme gradient boosting (XGBoost), random forest (RF), and neural network (NN)-to develop the models. The area under the receiver operating characteristic (AUC-ROC) curve, calibration curve, and variable importance were used to evaluate the models' performance.

RESULTS

During a median follow-up of 45.5 months, 179 patients (8.0%) experienced structural recurrence. The non-stimulated thyroglobulin, LN dissection, number of LNs dissected, lymph node metastasis ratio, N stage, comorbidity of hypertension, comorbidity of diabetes, body mass index, and low-density lipoprotein were used to develop the models. All models showed a greater AUC (AUC = 0.738 to 0.767) than did the ATA risk stratification (AUC = 0.620, DeLong test: P < 0.01). The SVM, XGBoost, and RF model showed greater sensitivity (0.568, 0.595, 0.676), specificity (0.903, 0.857, 0.784), accuracy (0.875, 0.835, 0.775), positive predictive value (PPV) (0.344, 0.272, 0.219), negative predictive value (NPV) (0.959, 0.959, 0.964), and F1 score (0.429, 0.373, 0.331) than did the ATA risk stratification (sensitivity = 0.432, specificity = 0.770, accuracy = 0.742, PPV = 0.144, NPV = 0.938, F1 score = 0.216). The RF model had generally consistent calibration compared with the other models. The Tg and the LNR were the top 2 important variables in all the models, the N stage was the top 5 important variables in all the models.

CONCLUSIONS

The RF model achieved the expected prediction performance with generally good discrimination, calibration and interpretability in this study. This study sheds light on the potential of ML approaches for improving the accuracy of risk stratification for PTC patients.

TRIAL REGISTRATION

Retrospectively registered at www.chictr.org.cn (trial registration number: ChiCTR2300075574, date of registration: 2023-09-08).

Construction and validation of a machine learning model for the diagnosis of juvenile idiopathic arthritis based on fecal microbiota

Purpose

Human gut microbiota has been shown to be significantly associated with various inflammatory diseases. Therefore, this study aimed to develop an excellent auxiliary tool for the diagnosis of juvenile idiopathic arthritis (JIA) based on fecal microbial biomarkers.

Method

The fecal metagenomic sequencing data associated with JIA were extracted from NCBI, and the sequencing data were transformed into the relative abundance of microorganisms by professional data cleaning (KneadData, Trimmomatic and Bowtie2) and comparison software (Kraken2 and Bracken). After that, the fecal microbes with high abundance were extracted for subsequent analysis. The extracted fecal microbes were further screened by least absolute shrinkage and selection operator (LASSO) regression, and the selected fecal microbe biomarkers were used for model training. In this study, we constructed six different machine learning (ML) models, and then selected the best model for constructing a JIA diagnostic tool by comparing the performance of the models based on a combined consideration of area under receiver operating characteristic curve (AUC), accuracy, specificity, F1 score, calibration curves and clinical decision curves. In addition, to further explain the model, Permutation Importance analysis and Shapley Additive Explanations (SHAP) were performed to understand the contribution of each biomarker in the prediction process.

Result

A total of 231 individuals were included in this study, including 203 JIA patients and Non-JIA individuals. In the analysis of diversity at the genus level, the alpha diversity represented by Shannon value was not significantly different between the two groups, while the belt diversity was slightly different. After selection by LASSO regression, 10 fecal microbe biomarkers were selected for model training. By comparing six different models, the XGB model showed the best performance, which average AUC, accuracy and F1 score were 0.976, 0.914 and 0.952, respectively, thus being used to construct the final JIA diagnosis model.

Conclusion

A JIA diagnosis model based on XGB algorithm was constructed with excellent performance, which may assist physicians in early detection of JIA patients and improve the prognosis of JIA patients.

Construction and validation of machine learning models for predicting distant metastases in newly diagnosed colorectal cancer patients: A large-scale and real-world cohort study

BACKGROUND

More accurate prediction of distant metastases (DM) in patients with colorectal cancer (CRC) would optimize individualized treatment and follow-up strategies. Multiple prediction models based on machine learning have been developed to assess the likelihood of developing DM.

METHODS

Clinicopathological features of patients with CRC were obtained from the National Cancer Center (NCC, China) and the Surveillance, Epidemiology, and End Results (SEER) database. The algorithms used to create the prediction models included random forest (RF), logistic regression, extreme gradient boosting, deep neural networks, and the K-Nearest Neighbor machine. The prediction models' performances were evaluated using receiver operating characteristic (ROC) curves.

RESULTS

In total, 200,958 patients, 3241 from NCC and 197,717 CRC from SEER were identified, of whom 21,736 (10.8%) developed DM. The machine-learning-based prediction models for DM were constructed with 12 features remaining after iterative filtering. The RF model performed the best, with areas under the ROC curve of 0.843, 0.793, and 0.806, respectively, on the training, test, and external validation sets. For the risk stratification analysis, the patients were separated into high-, middle-, and low-risk groups according to their risk scores. Patients in the high-risk group had the highest incidence of DM and the worst prognosis. Surgery, chemotherapy, and radiotherapy could significantly improve the prognosis of the high-risk and middle-risk groups, whereas the low-risk group only benefited from surgery and chemotherapy.

CONCLUSION

The RF-based model accurately predicted the likelihood of DM and identified patients with CRC in the high-risk group, providing guidance for personalized clinical decision-making.

New perspectives on cancer clinical research in the era of big data and machine learning

In the 21st century, the development of medical science has entered the era of big data, and machine learning has become an essential tool for mining medical big data. The establishment of the SEER database has provided a wealth of epidemiological data for cancer clinical research, and the number of studies based on SEER and machine learning has been growing in recent years. This article reviews recent research based on SEER and machine learning and finds that the current focus of such studies is primarily on the development and validation of models using machine learning algorithms, with the main directions being lymph node metastasis prediction, distant metastasis prediction, and prognosis-related research. Compared to traditional models, machine learning algorithms have the advantage of stronger adaptability, but also suffer from disadvantages such as overfitting and poor interpretability, which need to be weighed in practical applications. At present, machine learning algorithms, as the foundation of artificial intelligence, have just begun to emerge in the field of cancer clinical research. The future development of oncology will enter a more precise era of cancer research, characterized by larger data, higher dimensions, and more frequent information exchange. Machine learning is bound to shine brightly in this field.

Machine Learning as a Tool for Early Detection: A Focus on Late-Stage Colorectal Cancer across Socioeconomic Spectrums

PURPOSE

To assess the efficacy of various machine learning (ML) algorithms in predicting late-stage colorectal cancer (CRC) diagnoses against the backdrop of socio-economic and regional healthcare disparities.

METHODS

An innovative theoretical framework was developed to integrate individual- and census tract-level social determinants of health (SDOH) with sociodemographic factors. A comparative analysis of the ML models was conducted using key performance metrics such as AUC-ROC to evaluate their predictive accuracy. Spatio-temporal analysis was used to identify disparities in late-stage CRC diagnosis probabilities.

RESULTS

Gradient boosting emerged as the superior model, with the top predictors for late-stage CRC diagnosis being anatomic site, year of diagnosis, age, proximity to superfund sites, and primary payer. Spatio-temporal clusters highlighted geographic areas with a statistically significant high probability of late-stage diagnoses, emphasizing the need for targeted healthcare interventions.

CONCLUSIONS

This research underlines the potential of ML in enhancing the prognostic predictions in oncology, particularly in CRC. The gradient boosting model, with its robust performance, holds promise for deployment in healthcare systems to aid early detection and formulate localized cancer prevention strategies. The study's methodology demonstrates a significant step toward utilizing AI in public health to mitigate disparities and improve cancer care outcomes.

AMIGO2 is a pivotal therapeutic target related to M2 polarization of macrophages in pancreatic ductal adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) is a common kind of lethal cancer, with low early diagnostic rate and poor prognosis. In this study, we identified and verified the AMIGO2 with significant diagnostic and prognostic value in PDAC through LASSO regression combined with multiple machines learning methods, including RVM-RFE and Random Forest in TCGA and GEO datasets. The relevance between the expression of AMIGO2 and M2 polarization of macrophages was identified through pancancer, normal tissue, and cell lines data in TCGA, GTEx and CCLE datasets. The relevance between AMIGO2 and M2 polarization was then further identified in our local PDAC cohort. Finally, the role of AMIGO2 as cancer promoter and pivotal factor enrolled in M2 polarization was verified through siRNA transfection and M2 macrophages induction. These findings could facilitate diagnosis and treatment of PDAC. In addition, further research was deemed necessary on the deep mechanism between AMIGO2 and M2 polarization of macrophages in PDAC.

Machine learning for the prediction of postoperative nosocomial pulmonary infection in patients with spinal cord injury

PURPOSE

The purpose of this study was to establish the best prediction model for postoperative nosocomial pulmonary infection through machine learning (ML) and assist physicians to make accurate diagnosis and treatment decisions.

METHODS

Patients with spinal cord injury (SCI) who admitted to a general hospital between July 2014 and April 2022 were included in this study. The data were segmented according to the ratio of seven to three, 70% were randomly selected to train the model, and the other 30% were used for testing. We used LASSO regression to screen the variables, and the selected variables were used in the construction of six different ML models. Shapley additive explanations and permutation importance were used to explain the output of the ML models. Finally, sensitivity, specificity, accuracy and area under receiver operating characteristic curve (AUC) were used as the evaluation index of the model.

RESULTS

A total of 870 patients were enrolled in this study, of whom 98 (11.26%) developed pulmonary infection. Seven variables were used for ML model construction and multivariate logistic regression analysis. Among these variables, age, ASIA scale and tracheotomy were found to be the independent risk factors for postoperative nosocomial pulmonary infection in SCI patients. Meanwhile, the prediction model based on RF algorithm performed best in the training and test sets. (AUC = 0.721, accuracy = 0.664, sensitivity = 0.694, specificity = 0.656).

CONCLUSION

Age, ASIA scale and tracheotomy were the independent risk factors of postoperative nosocomial pulmonary infection in SCI. The prediction model based on RF algorithm had the best performance.

Bone metastasis detection method based on improving golden jackal optimization using whale optimization algorithm

This paper presents a machine learning-based technique for interpreting bone scintigraphy images, focusing on feature extraction and introducing a new feature selection method called GJOW. GJOW enhances the effectiveness of the golden jackal optimization (GJO) algorithm by integrating operators from the whale optimization algorithm (WOA). The technique's performance is evaluated through extensive experiments using 18 benchmark datasets and 581 bone scan images obtained from a gamma camera, including 362 abnormal and 219 normal cases. The results highlight the superior predictive effectiveness of the GJOW algorithm in bone metastasis detection, achieving an accuracy of 71.79% and specificity of 91.14%. The contributions of this study include the introduction of a new machine learning-based approach for detecting bone metastasis using gamma camera scans, leading to improved accuracy in identifying bone metastases. The findings have practical implications for early detection and intervention, potentially improving patient outcomes.

Artificial intelligence in the diagnosis of thyroid cancer: Recent advances and future directions

The diagnosis and management of thyroid cancer is fraught with challenges despite the advent of innovative diagnostic, surgical, and chemotherapeutic modalities. Challenges like inaccuracy in prognostication, uncertainty in cytopathological diagnosis, trouble in differentiating follicular neoplasms, intra-observer and inter-observer variability on ultrasound imaging preclude personalised treatment in thyroid cancer. Artificial intelligence (AI) is bringing a paradigm shift to the healthcare, powered by quick advancement of the analytic techniques. Several recent studies have shown remarkable progress in thyroid cancer diagnostics based on AI-assisted algorithms. Application of AI techniques in thyroid ultrasonography and cytopathology have shown remarkable impro-vement in sensitivity and specificity over the traditional diagnostic modalities. AI has also been explored in the development of treatment algorithms for indeterminate nodules and for prognostication in the patients with thyroid cancer. The benefits of high repeatability and straightforward implementation of AI in the management of thyroid cancer suggest that it holds promise for clinical application. Limited clinical experience and lack of prospective validation studies remain the biggest drawbacks. Developing verified and trustworthy algorithms after extensive testing and validation using prospective, multi-centre trials is crucial for the future use of AI in the pipeline of precision medicine in the management of thyroid cancer.

Risk factors and prognostic nomogram for patients with second primary cancers after lung cancer using classical statistics and machine learning

Previous studies have revealed an increased risk of secondary primary cancers (SPC) after lung cancer. The prognostic prediction models for SPC patients after lung cancer are particularly needed to guide screening. Therefore, we study retrospectively analyzed the Surveillance, Epidemiology, and End Results (SEER) database using classical statistics and machine learning to explore the risk factors and construct a novel overall survival (OS) prediction nomogram for patients with SPC after lung cancer. Data of patients with SPC after lung cancer, covering 2000 to 2016, were gathered from the SEER database. The incidence of SPC after lung cancer was calculated by Standardized incidence ratios (SIRs). Cox proportional hazards regression, machine learning (ML), Kaplan-Meier (KM) methods, and log-rank tests were conducted to identify the important prognostic factors for predicting OS. These significant prognostic factors were used for the development of an OS prediction nomogram. Totally, 10,487 SPC samples were randomly divided into training and validation cohorts (model construction and internal validation) from the SEER database. In the random forest (RF) and extreme gradient boosting (XGBoost) feature importance ranking models, age was the most important variable which was also reflected in the nomogram. And, the models that combined machine learning with cox proportional hazards had a better predictive performance than the model that only used cox proportional hazards (AUC = 0.762 in RF, AUC = 0.737 in XGBoost, AUC = 0.722 in COX). Calibration curves and decision curve analysis (DCA) curves also revealed that our nomogram has excellent clinical utility. The web-based dynamic nomogram calculator was accessible on https://httseer.shinyapps.io/DynNomapp/ . The prognosis characteristics of SPC following lung cancer were systematically reviewed. The dynamic nomogram we constructed can provide survival predictions to assist clinicians in making individualized decisions.

Artificial Intelligence in Thyroidology: A Narrative Review of the Current Applications, Associated Challenges, and Future Directions

Background: The use of artificial intelligence (AI) in health care has grown exponentially with the promise of facilitating biomedical research and enhancing diagnosis, treatment, monitoring, disease prevention, and health care delivery. We aim to examine the current state, limitations, and future directions of AI in thyroidology. Summary: AI has been explored in thyroidology since the 1990s, and currently, there is an increasing interest in applying AI to improve the care of patients with thyroid nodules (TNODs), thyroid cancer, and functional or autoimmune thyroid disease. These applications aim to automate processes, improve the accuracy and consistency of diagnosis, personalize treatment, decrease the burden for health care professionals, improve access to specialized care in areas lacking expertise, deepen the understanding of subtle pathophysiologic patterns, and accelerate the learning curve of less experienced clinicians. There are promising results for many of these applications. Yet, most are in the validation or early clinical evaluation stages. Only a few are currently adopted for risk stratification of TNODs by ultrasound and determination of the malignant nature of indeterminate TNODs by molecular testing. Challenges of the currently available AI applications include the lack of prospective and multicenter validations and utility studies, small and low diversity of training data sets, differences in data sources, lack of explainability, unclear clinical impact, inadequate stakeholder engagement, and inability to use outside of the research setting, which might limit the value of their future adoption. Conclusions: AI has the potential to improve many aspects of thyroidology; however, addressing the limitations affecting the suitability of AI interventions in thyroidology is a prerequisite to ensure that AI provides added value for patients with thyroid disease.

Osseous Metastases in Thyroid Cancer: Unveiling Risk Factors, Disease Outcomes, and Treatment Impact

Bone is the second most common site of metastasis in patients with thyroid cancer (TC) and dramatically impacts overall survival and quality of life with no definitive cure, yet there is no extensive study of the demographic and clinical risk factors in the recent literature. Data regarding 120,754 TC patients with bone metastasis were retrieved from the Surveillance, Epidemiology, and End Results (SEER) database. Univariate and multivariate analyses were used to identify the risk factors of bone metastasis occurring in various histologies of TC. Cox regression was performed to analyze the influence of bone metastasis on overall survival. Hazard ratios were computed to analyze the association between bone metastasis and the primary outcomes. Of the 120,754 records collected from the SEER database from 2000 to 2019, 976 (0.8%) presented with bone metastasis, with occurrence being the greatest in patients of age ≥ 55 years (OR = 5.63, 95%CI = 4.72-6.71), males (OR = 2.60, 95%CI = 2.27-2.97), Blacks (OR = 2.38, 95%CI = 1.95-2.9) and Asian or Pacific Islanders (OR = 1.90, 95%CI = 1.58-2.27), and single marital status. TC patients presenting with bone metastasis (HR = 2.78, 95%CI = 2.34-3.3) or concurrent bone and brain metastases (HR = 1.62, 95%CI = 1.03-2.55) had a higher mortality risk. Older age, gender, race, and single marital status were associated with bone metastasis and poorer prognosis in TC patients at initial diagnosis. Understanding such risk factors can potentially assist clinicians in making early diagnoses and personalized treatment plans, as well as researchers in developing more therapeutic protocols.

Development of a machine learning algorithm to predict complications of total laparoscopic anterior resection and natural orifice specimen extraction surgery in rectal cancer

BACKGROUND

Total laparoscopic anterior resection (tLAR) and natural orifice specimen extraction surgery (NOSES) has been widely adopted in the treatment of rectal cancer (RC). However, no study has been performed to predict the short-term outcomes of tLAR using machine learning algorithms to analyze a national cohort.

METHODS

Data from consecutive RC patients who underwent tLAR were collected from the China NOSES Database (CNDB). The random forest (RF), extreme gradient boosting (XGBoost), support vector machine (SVM), deep neural network (DNN), logistic regression (LR) and K-nearest neighbor (KNN) algorithms were used to develop risk models to predict short-term complications of tLAR. The area under the receiver operating characteristic curve (AUROC), Gini coefficient, specificity and sensitivity were calculated to assess the performance of each risk model. The selected factors from the models were evaluated by relative importance.

RESULTS

A total of 4313 RC patients were identified, and 667 patients (15.5%) developed postoperative complications. The machine learning model of XGBoost showed more promising results in the prediction of complication than other models (AUROC 0.90, P < 0.001). The performance was similar when internal and external validation was used. In the XGBoost model, the top four influential factors were the distance from the lower edge of the tumor to the anus, age at diagnosis, surgical time and comorbidities. In risk stratification analysis, the rate of postoperative complications in the high-risk group was significantly higher than in the medium- and low-risk groups (P < 0.001).

CONCLUSION

The machine learning model shows potential benefits in predicting the risk of complications in RC patients after tLAR. This novel approach can provide reliable individual information for surgical treatment recommendations.

Predicting five-year mortality in soft-tissue sarcoma patients

Aims

The aim of this study was to identify factors associated with five-year cancer-related mortality in patients with limb and trunk soft-tissue sarcoma (STS) and develop and validate machine learning algorithms in order to predict five-year cancer-related mortality in these patients.

Methods

Demographic, clinicopathological, and treatment variables of limb and trunk STS patients in the Surveillance, Epidemiology, and End Results Program (SEER) database from 2004 to 2017 were analyzed. Multivariable logistic regression was used to determine factors significantly associated with five-year cancer-related mortality. Various machine learning models were developed and compared using area under the curve (AUC), calibration, and decision curve analysis. The model that performed best on the SEER testing data was further assessed to determine the variables most important in its predictive capacity. This model was externally validated using our institutional dataset.

Results

A total of 13,646 patients with STS from the SEER database were included, of whom 35.9% experienced five-year cancer-related mortality. The random forest model performed the best overall and identified tumour size as the most important variable when predicting mortality in patients with STS, followed by M stage, histological subtype, age, and surgical excision. Each variable was significant in logistic regression. External validation yielded an AUC of 0.752.

Conclusion

This study identified clinically important variables associated with five-year cancer-related mortality in patients with limb and trunk STS, and developed a predictive model that demonstrated good accuracy and predictability. Orthopaedic oncologists may use these findings to further risk-stratify their patients and recommend an optimal course of treatment.

Prediction of Non-Home Discharge Following Total Hip Arthroplasty in Geriatric Patients

Introduction

The majority of total hip arthroplasty (THA) patients are discharged home postoperatively, however, many still require continued medical care. We aimed to identify important characteristics that predict nonhome discharge in geriatric patients undergoing THA using machine learning. We hypothesize that our analyses will identify variables associated with decreased functional status and overall health to be predictive of non-home discharge.

Materials and Methods

Elective, unilateral, THA patients above 65 years of age were isolated in the NSQIP database from 2018-2020. Demographic, pre-operative, and intraoperative variables were analyzed. After splitting the data into training (75%) and validation (25%) data sets, various machine learning models were used to predict non-home discharge. The model with the best area under the curve (AUC) was further assessed to identify the most important variables.

Results

In total, 19,840 geriatric patients undergoing THA were included in the final analyses, of which 5194 (26.2%) were discharged to a non-home setting. The RF model performed the best and identified age above 78 years (OR: 1.08 [1.07, 1.09], P < .0001), as the most important variable when predicting non-home discharge in geriatric patients with THA, followed by severe American Society of Anesthesiologists grade (OR: 1.94 [1.80, 2.10], P < .0001), operation time (OR: 1.01 [1.00, 1.02], P < .0001), anemia (OR: 2.20 [1.87, 2.58], P < .0001), and general anesthesia (OR: 1.64 [1.52, 1.79], P < .0001). Each of these variables was also significant in MLR analysis. The RF model displayed good discrimination with AUC = .831.

Discussion

The RF model revealed clinically important variables for assessing discharge disposition in geriatric patients undergoing THA, with the five most important factors being older age, severe ASA grade, longer operation time, anemia, and general anesthesia.

Conclusions

With the rising emphasis on patient-centered care, incorporating models such as these may allow for preoperative risk factor mitigation and reductions in healthcare expenditure.

Development of rapid and effective risk prediction models for stroke in the Chinese population: a cross-sectional study

OBJECTIVES

The purpose of this study was to use easily obtained and directly observable clinical features to establish predictive models to identify patients at increased risk of stroke.

SETTING AND PARTICIPANTS

A total of 46 240 valid records were obtained from 8 research centres and 14 communities in Jiangxi province, China, between February and September 2018.

PRIMARY AND SECONDARY OUTCOME MEASURES

The area under the receiver operating characteristic curve (AUC), sensitivity, specificity and accuracy were calculated to test the performance of the five models (logistic regression (LR), random forest (RF), decision tree (DT), extreme gradient boosting (XGBoost) and gradient boosting DT). The calibration curve was used to show calibration performance.

RESULTS

The results indicated that XGBoost (AUC: 0.924, accuracy: 0.873, sensitivity: 0.776, specificity: 0.916) and RF (AUC: 0.924, accuracy: 0.872, sensitivity: 0.778, specificity: 0.913) demonstrated excellent performance in predicting stroke. Physical inactivity, hypertension, meat-based diet and high salt intake were important prediction features of stroke.

CONCLUSION

The five machine learning models all had good predictive and discriminatory performance for stroke. The performance of RF and XGBoost was slightly better than that of LR, which was easier to interpret and less prone to overfitting. This work provides a rapid and accurate tool for stroke risk assessment, which can help to improve the efficiency of stroke screening medical services and the management of high-risk groups.

Deep learning of bone metastasis in small cell lung cancer: A large sample-based study

Introduction

Bone is a common metastatic site for small cell lung cancer (SCLC). Bone metastasis (BM) in patients have are known to show poor prognostic outcomes. We explored the epidemiological characteristics of BM in SCLC patients and create a new deep learning model to predict outcomes for cancer-specific survival (CSS) and overall survival (OS).

Materials and Methods

Data for SCLC patients diagnosed with or without BM from 2010 to 2016 were retrieved from the Surveillance, Epidemiology, and End Results (SEER) database. Univariate and multivariate Cox proportional hazards regression models were used to evaluate the effects of prognostic variables on OS and CSS. Through integration of these variables, nomograms were created for the prediction of CSS and OS rates at 3-month,6- month,and 12-month. Harrell's coordination index, calibration curves,and time- dependent ROC curves were used to assess the nomograms' accuracy. Decision tree analysis was used to evaluate the clinical application value of the established nomogram.

Results

In this study, 4201 patients were enrolled. Male sex, tumor size 25 but <10, brain and liver metastases, as well as chemotherapy were associated with a high risk for BM. Tumor size, Age, N stage, gender, liver metastasis, radiotherapy as well as chemotherapy were shown to be prognostic variables for OS, and the prognostic variables for CSS were added to the tumor number in addition. Based on these results, nomograms for CSS and OS were established separately. Internal as well as external validation showed that the C-index, calibration cuurve and DCA had good constructive correction effect and clinical application value. Decision tree analysis further confirmed the prognostic factors of OS and CSS.

Discussion

The nomogram and decision tree models developed in this study effectively guided treatment decisions for SCLC patients with BM. The creation of prediction models for BM SCLC patients may be facilitated by deep learning models.

Peripheral and tumor-infiltrating immune cells are correlated with patient outcomes in ovarian cancer

OBJECTIVE

At present, there is still a lack of reliable biomarkers for ovarian cancer (OC) to guide prognosis prediction and accurately evaluate the dominant population of immunotherapy. In recent years, the relationship between peripheral blood markers and tumor-infiltrating immune cells (TICs) with cancer has attracted much attention. However, the relationship between the survival of OC patients and intratumoral- or extratumoral-associated immune cells remains controversial.

METHODS

In this study, four machine-learning algorithms were used to predict overall survival in OC patients based on peripheral blood indicators. To further screen out immune-related gene and molecular targets, we systematically explored the correlation between TICs and OC patient survival based on The Cancer Genome Atlas database. Using the TICs score method, patients were divided into a low immune infiltrating cell group and a high immune infiltrating cell group.

RESULTS

The results showed that there was a significant statistical significance between the peripheral blood indicators and the survival prognosis of OC patients. Survival analysis showed that TICs play a crucial role in the survival of OC patients. Four core genes, CXCL9, CD79A, MS4A1, and MZB1, were identified by cross-PPI and COX regression analysis. Further analysis found that these genes were significantly associated with both TICs and survival in OC patients.

CONCLUSIONS

These results suggest that both peripheral blood markers and TICs can be used as prognostic predictors in patients with OC, and CXCL9, CD79A, MS4A1, and MZB1 may be potential therapeutic targets for OC immunotherapy.

Blood Biomarkers Panels for Screening of Colorectal Cancer and Adenoma on a Machine Learning-Assisted Detection Platform

OBJECTIVE

A mini-invasive and good-compliance program is critical to broaden colorectal cancer (CRC) screening and reduce CRC-related mortality. Blood testing combined with imaging examination has been proved to be feasible on screen for multicancer and guide intervention. The study aims to construct a machine learning-assisted detection platform with available multi-targets for CRC and colorectal adenoma (CRA) screening.

METHODS

This was a retrospective study that the blood test data from 204 CRCs, 384 CRAs, and 229 healthy controls was extracted. The classified models were constructed with 4 machine learning (ML) algorithms including support vector machine (SVM), random forest (RF), decision tree (DT), and eXtreme Gradient Boosting (XGB) based on the candidate biomarkers. The importance index was used by SHapely Adaptive exPlanations (SHAP) analysis to identify the dominant characteristics. The performance of classified models was evaluated. The most dominating features from the proposed panel were developed by logistic regression (LR) for identification CRC from control.

RESULTS

The candidate biomarkers consisted of 26 multi-targets panel including CEA, AFP, and so on. Among the 4 models, the SVM classifier for CRA yields the best predictive performance (the area under the receiver operating curve, AUC: .925, sensitivity: .904, and specificity: .771). As for CRC classification, the RF model with 26 candidate biomarkers provided the best predictive parameters (AUC: .941, sensitivity: .902, and specificity: .912). Compared with CEA and CA199, the predictive performance was significantly improved. The streamlined model with 6 biomarkers for CRC also obtained a good performance (AUC: .946, sensitivity: .885, and specificity: .913).

CONCLUSIONS

The predictive models consisting of 26 multi-targets panel would be used as a non-invasive, economical, and effective risk stratification platform, which was expected to be applied for auxiliary screening of CRA and CRC in clinical practice.

A practical dynamic nomogram model for predicting bone metastasis in patients with thyroid cancer

PURPOSE

The aim of this study was to established a dynamic nomogram for assessing the risk of bone metastasis in patients with thyroid cancer (TC) and assist physicians to make accurate clinical decisions.

METHODS

The clinical data of patients with TC admitted to the First Affiliated hospital of Nanchang University from January 2006 to November 2016 were included in this study. Demographic and clinicopathological parameters of all patients at primary diagnosis were analyzed. Univariate and multivariate logistic regression analysis was applied to build a predictive model incorporating parameters. The discrimination, calibration, and clinical usefulness of the nomogram were evaluated using the C-index, ROC curve, calibration plot, and decision curve analysis. Internal validation was evaluated using the bootstrapping method.

RESULTS

A total of 565 patients were enrolled in this study, of whom 25 (4.21%) developed bone metastases. Based on logistic regression analysis, age (OR=1.040, P=0.019), hemoglobin (HB) (OR=0.947, P<0.001) and alkaline phosphatase (ALP) (OR=1.006, P=0.002) levels were used to construct the nomogram. The model exhibited good discrimination, with a C-index of 0.825 and good calibration. A C-index value of 0.815 was achieved on interval validation analysis. Decision curve analysis showed that the nomogram was clinically useful when intervention was decided at a bone metastases possibility threshold of 1%.

CONCLUSIONS

This dynamic nomogram, with relatively good accuracy, incorporating age, HB, and ALP, could be conveniently used to facilitate the prediction of bone metastasis risk in patients with TC.

Bone metastases in newly diagnosed patients with thyroid cancer: A large population-based cohort study

Background

Population-based estimates of the incidence and prognosis of bone metastases (BM) stratified by histologic subtype at diagnosis of thyroid cancer are limited.

Methods

Using multivariable logistic and Cox regression analyses, we identified risk factors for BM and investigated the prognostic survival of BM patients between 2010 and 2015 via the Surveillance, Epidemiology, and End Results (SEER) database.

Results

Among 64,083 eligible patients, a total of 347 patients with BM at the time of diagnosis were identified, representing 0.5% of the entire cohort but 32.4% of the subset with metastases. BM incidence was highest (11.6%) in anaplastic thyroid cancer (ATC), which, nevertheless, was highest (61.5%) in follicular thyroid cancer (FTC) among the subset with metastases. The median overall survival among BM patients was 40.0 months, and 1-, 3-, and 5-year survival rates were 65.2%, 51.3%, and 38.7%, respectively. Compared with papillary thyroid cancer (PTC), FTC (aOR, 6.33; 95% CI, 4.72–8.48), medullary thyroid cancer (MTC) (aOR, 6.04, 95% CI, 4.09–8.92), and ATC (aOR, 6.21; 95% CI, 4.20–9.18) significantly increased the risk of developing BM. However, only ATC (aHR, 6.07; 95% CI, 3.83–9.60) was independently associated with worse survival in multivariable analysis. Additionally, patients with BM alone (56.5%) displayed the longest median survival (66.0 months), compared with those complicated with one extraskeletal metastatic site (lung, brain, or liver) (35.2%; 14.0 months) and two or three sites (8.3%; 6.0 months). The former 5-year overall survival rate was 52.6%, which, however, drastically declined to 23.0% in patients with one extraskeletal metastatic site and 9.1% with two or three sites.

Conclusion

Closer bone surveillance should be required for patients with FTC, MTC, and ATC, and extraskeletal metastases at initial diagnosis frequently predict a poorer prognosis.

Using Machine Learning Methods to Predict Bone Metastases in Breast Infiltrating Ductal Carcinoma Patients

Breast cancer (BC) was the most common malignant tumor in women, and breast infiltrating ductal carcinoma (IDC) accounted for about 80% of all BC cases. BC patients who had bone metastases (BM) were more likely to have poor prognosis and bad quality of life, and earlier attention to patients at a high risk of BM was important. This study aimed to develop a predictive model based on machine learning to predict risk of BM in patients with IDC. Six different machine learning algorithms, including Logistic regression (LR), Naive Bayes classifiers (NBC), Decision tree (DT), Random Forest (RF), Gradient Boosting Machine (GBM), and Extreme gradient boosting (XGB), were used to build prediction models. The XGB model offered the best predictive performance among these 6 models in internal and external validation sets (AUC: 0.888, accuracy: 0.803, sensitivity: 0.801, and specificity: 0.837). Finally, an XGB model-based web predictor was developed to predict risk of BM in IDC patients, which may help physicians make personalized clinical decisions and treatment plans for IDC patients.

A Predictive Model for the 10-year Overall Survival Status of Patients With Distant Metastases From Differentiated Thyroid Cancer Using XGBoost Algorithm-A Population-Based Analysis

Purpose: To explore clinical and non-clinical characteristics affecting the prognosis of patients with differentiated thyroid cancer with distant metastasis (DTCDM) and establish an accurate overall survival (OS) prognostic model.

Patients and methods: Study subjects and related information were obtained from the National Cancer Institute’s surveillance, epidemiology, and results database (SEER). Kaplan‐Meier analysis, log-rank test, and univariate and multivariate Cox analysis were used to screen for factors influencing the OS of patients with DTCDM. Nine variables were introduced to build a machine learning (ML) model, receiver operating characteristic (ROC) was used to evaluate the recognition ability of the model, calibration plots were used to obtain prediction accuracy, and decision curve analysis (DCA) was used to estimate clinical benefit.

Results: After applying the inclusion and exclusion criteria, a total of 3,060 patients with DTCDM were included in the survival analysis from 2004 to 2017. A machine learning prediction model was developed with nine variables: age at diagnosis, gender, race, tumor size, histology, regional lymph node metastasis, primary site surgery, radiotherapy, and chemotherapy. After excluding patients who survived <120 months, variables were sub-coded and machine learning was used to model OS prognosis in patients with DTCDM. Patients 6–50 years of age had the highest scores in the model. Other variables with high scores included small tumor size, male sex, and age 51–76. The AUC and calibration curves confirm that the XGBoost model has good performance. DCA shows that our model can be used to support clinical decision-making in a 10-years overall survival model.

Conclusion: An artificial intelligence model was constructed using the XGBoost algorithms to predict the 10-years overall survival rate of patients with DTCDM. After model validation and evaluation, the model had good discriminative ability and high clinical value. This model could serve as a clinical tool to help inform treatment decisions for patients with DTCDM.

Risk factors associated with skeletal-related events following discontinuation of denosumab treatment among patients with bone metastases from solid tumors: A real-world machine learning approach

•

This study investigated SRE risk factors after densomuab treatment discontinuation.

•

An unbiased machine learning approach was developed to evaluate >60 variables.

•

Prior SREs and short denosumab treatment duration were primary risk factors.

•

The results can guide denosumab persistence decisions and improve patient outcomes.

Background

Clinical practice guidelines recommend the use of bone-targeting agents for preventing skeletal-related events (SREs) among patients with bone metastases from solid tumors. The anti-RANKL monoclonal antibody denosumab is approved for the prevention of SREs in patients with bone metastases from solid tumors. However, real-world data are lacking on the impact of individual risk factors for SREs, specifically in the context of denosumab discontinuation.

Purpose

We aim to identify risk factors associated with SRE incidence following denosumab discontinuation using a machine learning approach to help profile patients at a higher risk of developing SREs following discontinuation of denosumab treatment.

Methods

Using the Optum PanTher Electronic Health Record repository, patients diagnosed with incident bone metastases from primary solid tumors between January 1, 2007, and September 1, 2019, were evaluated for inclusion in the study. Eligible patients received ≥ 2 consecutive 120 mg denosumab doses on a 4-week (± 14 days) schedule with a minimum follow-up of ≥ 1 year after the last denosumab dose, or an SRE occurring between days 84 and 365 after denosumab discontinuation. Extreme gradient boosting was used to develop an SRE risk prediction model evaluated on a test dataset. Multiple variables associated with patient demographics, comorbidities, laboratory values, treatments, and denosumab exposures were examined as potential factors for SRE risk using Shapley Additive Explanations (SHAP). Univariate analyses on risk factors with the highest importance from pooled and tumor-specific models were also conducted.

Results

A total of 1,414 adult cancer patients (breast: 40%, prostate: 30%, lung: 13%, other: 17%) were eligible, of whom 1,133 (80%) were assigned to model training and 281 (20%) to model evaluation. The median age at inclusion was 67 (range, 19–89) years with a median duration of denosumab treatment of 253 (range, 88–2,726) days; 490 (35%) patients experienced ≥ 1 SRE 83 days after denosumab discontinuation. Meaningful model performance was evaluated by an area under the receiver operating curve score of 77% and an F1 score of 62%; model precision was 60%, with 63% sensitivity and 78% specificity. SHAP identified several significant factors for the tumor-agnostic and tumor-specific models that predicted an increased SRE risk following denosumab discontinuation, including prior SREs, shorter denosumab treatment duration, ≥ 4 clinic visits per month with at least one hospitalization (all-cause) event from the baseline period up to discontinuation of denosumab, younger age at bone metastasis, shorter time to denosumab initiation from bone metastasis, and prostate cancer.

Conclusion

This analysis showed a higher cumulative number of SREs, prior SREs relative to denosumab initiation, a higher number of hospital visits, and a shorter denosumab treatment duration as significant factors that are associated with an increased SRE risk after discontinuation of denosumab, in both the tumor-agnostic and tumor-specific models. Our machine learning approach to SRE risk factor identification reinforces treatment guidance on the persistent use of denosumab and has the potential to help clinicians better assess a patient’s need to continue denosumab treatment and improve patient outcomes.

Ten Years After SINS: Role of Surgery and Radiotherapy in the Management of Patients With Vertebral Metastases

The objective of the different types of treatments for a spinal metastasis is to provide the best oncological and functional result with the least aggressive side effects. Initially created in 2010 to help clinicians in the management of vertebral metastases, the Spine Instability Neoplastic Score (SINS) has quickly found its place in the decision making and the treatment of patients with metastatic spinal disease. Here we conduct a review of the literature describing the different changes that occurred with the SINS score in the last ten years. After a brief presentation of the spinal metastases’ distribution, with or without spinal cord compression, we present the utility of SINS in the radiological diagnosis and extension of the disease, in addition to its limits, especially for scores ranging between 7 and 12. We take this opportunity to expose the latest advances in surgery and radiotherapy concerning spinal metastases, as well as in palliative care and pain control. We also discuss the reliability of SINS amongst radiologists, radiation oncologists, spine surgeons and spine surgery trainees. Finally, we will present the new SINS-derived predictive scores, biomarkers and artificial intelligence algorithms that allow a multidisciplinary approach for the management of spinal metastases.

Application of Machine Learning Algorithms to Predict Lymph Node Metastasis in Early Gastric Cancer

Objective

This study aimed to establish the best early gastric cancer lymph node metastasis (LNM) prediction model through machine learning (ML) to better guide clinical diagnosis and treatment decisions.

Methods

We screened gastric cancer patients with T1a and T1b stages from 2010 to 2015 in the Surveillance, Epidemiology and End Results (SEER) database and collected the clinicopathological data of patients with early gastric cancer who were treated with surgery at the Second Affiliated Hospital of Nanchang University from January 2014 to December 2016. At the same time, we applied 7 ML algorithms—the generalized linear model (GLM), RPART, random forest (RF), gradient boosting machine (GBM), support vector machine (SVM), regularized dual averaging (RDA), and the neural network (NNET)—and combined them with patient pathological information to develop the best prediction model for early gastric cancer lymph node metastasis. Among the SEER set, 80% were randomly selected to train the models, while the remaining 20% were used for testing. The data from the Second Affiliated Hospital were considered as the external verification set. Finally, we used the AUROC, F1-score value, sensitivity, and specificity to evaluate the performance of the model.

Results

The tumour size, tumour grade, and depth of tumour invasion were independent risk factors for early gastric cancer LNM. Comprehensive comparison of the prediction model performance of the training set and test set showed that the RDA model had the best prediction performance (F1-score = 0.773; AUROC = 0.742). The AUROC of the external validation set was 0.73.

Conclusions

Tumour size, tumour grade, and depth of tumour invasion were independent risk factors for early gastric cancer LNM. ML predicted LNM risk more accurately, and the RDA model had the best predictive performance and could better guide clinical diagnosis and treatment decisions.

A Proposed Heterogeneous Ensemble Algorithm Model for Predicting Central Lymph Node Metastasis in Papillary Thyroid Cancer

Purpose

To develop a heterogeneous ensemble algorithm model to precisely predict central lymph node metastasis (CLNM), which can provide a reference value on controversial topics of performing prophylactic central lymph node dissection for patients with papillary thyroid cancer (PTC).

Methods

The study included patients with PTC who underwent an initial thyroid resection in a single-center medical institution between January 2014 and December 2018. A total of 18 variables, including clinical features and ultrasound (US) features, were used in the univariate analysis, multivariate analysis, and feature selection and were also used to develop a heterogeneous ensemble model based on five basic machine learning models, including extreme gradient boosting, k-nearest neighbors, random forest, gradient boosting, and AdaBoost. Moreover, a partial dependent plot was used to explain the heterogeneous ensemble model.

Results

The area under the receiver operating characteristic curve of the heterogeneous ensemble algorithm model was 0.67, which is significantly better than that of the basic machine models in predicting CLNM. All machine learning models performed better than US. Based on multivariate analysis and receiver operating characteristic curve analysis, age ≤33 years, tumor size ≥0.8 cm, US-suspected CLNM, and microcalcification were risk factors for CLNM, and anti-thyroid peroxidase antibody and serum thyroglobulin levels were favorable factors for CLNM.

Conclusion

The proposed heterogeneous ensemble algorithm model may be optimal tool to predict CLNM by integrating clinical and US features.

Application of Machine Learning Techniques to Predict Bone Metastasis in Patients with Prostate Cancer

Objective

This study aimed to develop and validate a machine learning model for predicting bone metastases (BM) in prostate cancer (PCa) patients.

Methods

Demographic and clinicopathologic variables of PCa patients in the Surveillance, Epidemiology and End Results (SEER) database from 2010 to 2017 were retrospectively analyzed. We used six different machine learning algorithms, including Decision tree (DT), Random forest (RF), Multilayer Perceptron (MLP), Logistic regression (LR), Naive Bayes classifiers (NBC), and eXtreme gradient boosting (XGB), to build prediction models. External validation using data from 644 PCa patients of the First Affiliated Hospital of Nanchang University from 2010 to 2016. The performance of the models was evaluated using the area under receiver operating characteristic curve (AUC), accuracy score, sensitivity (recall rate) and specificity. A web predictor was developed based on the best performance model.

Results

A total of 207,137 PCa patients from SEER were included in this study. Of whom, 6725 (3.25%) developed BM. Gleason score, Prostate-specific antigen (PSA) value, T, N stage and age were found to be the risk factors of BM. The XGB model offered the best predictive performance among these 6 models (AUC: 0.962, accuracy: 0.884, sensitivity (recall rate): 0.906, and specificity: 0.879). An XGB model-based web predictor was developed to predict BM in PCa patients.

Conclusion

This study developed a machine learning model and a web predictor for predicting the risk of BM in PCa patients, which may help physicians make personalized clinical decisions and treatment strategy for patients.