A review of statistical and machine learning methods for modeling cancer risk using structured clinical data

Development and validation of a risk scoring system to identify patients with lupus nephritis in electronic health record data

OBJECTIVE

Accurate identification of lupus nephritis (LN) cases is essential for patient management, research and public health initiatives. However, LN diagnosis codes in electronic health records (EHRs) are underused, hindering efficient identification. We investigated the current performance of International Classification of Diseases (ICD) codes, 9th and 10th editions (ICD9/10), for identifying prevalent LN, and developed scoring systems to increase identification of LN that are adaptable to settings with and without LN ICD codes.

METHODS

Training and test sets derived from EHR data from a large health system. An external set comprised data from the EHR of a second large health system. Adults with ICD9/10 codes for SLE were included. LN cases were ascertained through manual chart reviews conducted by rheumatologists. Two definitions of LN were used: strict (definite LN) and inclusive (definite, potential or diagnostic uncertainty). Gradient boosting models including structured EHR fields were used for predictor selection. Two logistic regression-based scoring systems were developed ('LN-Code' included LN ICD codes and 'LN-No Code' did not), calibrated and validated using standard performance metrics.

RESULTS

A total of 4152 patients from University of California San Francisco Medical Center and 370 patients from Zuckerberg San Francisco General Hospital and Trauma Center met the eligibility criteria. Mean age was 50 years, 87% were female. LN diagnosis codes demonstrated low sensitivity (43-73%) but high specificity (92-97%). LN-Code achieved an area under the curve (AUC) of 0.93 and a sensitivity of 0.88 for identifying LN using the inclusive definition. LN-No Code reached an AUC of 0.91 and a sensitivity of 0.95 (0.97 for the strict definition). Both scoring systems had good external validity, calibration and performance across racial and ethnic groups.

CONCLUSIONS

This study quantified the underutilisation of LN diagnosis codes in EHRs and introduced two adaptable scoring systems to enhance LN identification. Further validation in diverse healthcare settings is essential to ensure their broader applicability.

Nomogram to predict risk of neonatal mortality among preterm neonates admitted with sepsis at University of Gondar Comprehensive Specialized Hospital: risk prediction model development and validation

BACKGROUND

Mortality in premature neonates is a global public health problem. In developing countries, nearly 50% of preterm births ends with death. Sepsis is one of the major causes of death in preterm neonates. Risk prediction model for mortality in preterm septic neonates helps for directing the decision making process made by clinicians.

OBJECTIVE

We aimed to develop and validate nomogram for the prediction of neonatal mortality. Nomograms are tools which assist the clinical decision making process through early estimation of risks prompting early interventions.

METHODS

A three year retrospective follow up study was conducted at University of Gondar Comprehensive Specialized Hospital and a total of 603 preterm neonates with sepsis were included. Data was collected using KoboCollect and analyzed using STATA version 16 and R version 4.2.1. Lasso regression was used to select the most potent predictors and to minimize the problem of overfitting. Nomogram was developed using multivariable binary logistic regression analysis. Model performance was evaluated using discrimination and calibration. Internal model validation was done using bootstrapping. Net benefit of the nomogram was assessed through decision curve analysis (DCA) to assess the clinical relevance of the model.

RESULT

The nomogram was developed using nine predictors: gestational age, maternal history of premature rupture of membrane, hypoglycemia, respiratory distress syndrome, perinatal asphyxia, necrotizing enterocolitis, total bilirubin, platelet count and kangaroo-mother care. The model had discriminatory power of 96.7% (95% CI: 95.6, 97.9) and P-value of 0.165 in the calibration test before and after internal validation with brier score of 0.07. Based on the net benefit analysis the nomogram was found better than treat all and treat none conditions.

CONCLUSION

The developed nomogram can be used for individualized mortality risk prediction with excellent performance, better net benefit and have been found to be useful in clinical practice with contribution in preterm neonatal mortality reduction by giving better emphasis for those at high risk.

A Machine-Learning Approach to Assess Factors Associated With Hospitalization of Children and Youths in Psychiatric Crisis

OBJECTIVE

The authors used a machine-learning approach to model clinician decision making regarding psychiatric hospitalization of children and youths in crisis and to identify factors associated with the decision to hospitalize.

METHODS

Data consisted of 4,786 mobile crisis response team assessments of children and youths, ages 4.0-19.5 years (mean±SD=14.0±2.7 years, 56% female), in Nevada. The sample assessments were split into training and testing data sets. A random-forest machine-learning algorithm was used to identify variables related to the decision to hospitalize a child or youth after the crisis assessment. Results from the training sample were externally validated in the testing sample.

RESULTS

The random-forest model had good performance (area under the curve training sample=0.91, testing sample=0.92). Variables found to be important in the decision to hospitalize a child or youth were acute suicidality, followed by poor judgment or decision making, danger to others, impulsivity, runaway behavior, other risky behaviors, nonsuicidal self-injury, psychotic or depressive symptoms, sleep problems, oppositional behavior, poor functioning at home or with peers, depressive or schizophrenia spectrum disorders, and age.

CONCLUSIONS

In crisis settings, clinicians were found to mostly focus on acute factors that increased risk for danger to self or others (e.g., suicidality, poor judgment), current psychiatric symptoms (e.g., psychotic symptoms), and functioning (e.g., poor home functioning, problems with peer relationships) when deciding whether to hospitalize or stabilize a child or youth. To reduce psychiatric hospitalization, community-based services should target interventions to address these important factors associated with the need for a higher level of care among youths in psychiatric crisis.

A Machine Learning Approach to Assess Patients with Deep Neck Infection Progression to Descending Mediastinitis: Preliminary Results

BACKGROUND

Deep neck infection (DNI) is a serious infectious disease, and descending mediastinitis is a fatal infection of the mediastinum. However, no study has applied artificial intelligence to assess progression to descending mediastinitis in DNI patients. Thus, we developed a model to assess the possible progression of DNI to descending mediastinitis.

METHODS

Between August 2017 and December 2022, 380 patients with DNI were enrolled; 75% of patients (n = 285) were assigned to the training group for validation, whereas the remaining 25% (n = 95) were assigned to the test group to determine the accuracy. The patients' clinical and computed tomography (CT) parameters were analyzed via the k-nearest neighbor method. The predicted and actual progression of DNI patients to descending mediastinitis were compared.

RESULTS

In the training and test groups, there was no statistical significance (all p > 0.05) noted at clinical variables (age, gender, chief complaint period, white blood cells, C-reactive protein, diabetes mellitus, and blood sugar), deep neck space (parapharyngeal, submandibular, retropharyngeal, and multiple spaces involved, ≥3), tracheostomy performance, imaging parameters (maximum diameter of abscess and nearest distance from abscess to level of sternum notch), or progression to mediastinitis. The model had a predictive accuracy of 82.11% (78/95 patients), with sensitivity and specificity of 41.67% and 87.95%, respectively.

CONCLUSIONS

Our model can assess the progression of DNI to descending mediastinitis depending on clinical and imaging parameters. It can be used to identify DNI patients who will benefit from prompt treatment.

Synergy between imputed genetic pathway and clinical information for predicting recurrence in early stage non-small cell lung cancer

OBJECTIVE

Lung cancer exhibits unpredictable recurrence in low-stage tumors and variable responses to different therapeutic interventions. Predicting relapse in early-stage lung cancer can facilitate precision medicine and improve patient survivability. While existing machine learning models rely on clinical data, incorporating genomic information could enhance their efficiency. This study aims to impute and integrate specific types of genomic data with clinical data to improve the accuracy of machine learning models for predicting relapse in early-stage, non-small cell lung cancer patients.

METHODS

The study utilized a publicly available TCGA lung cancer cohort and imputed genetic pathway scores into the Spanish Lung Cancer Group (SLCG) data, specifically in 1348 early-stage patients. Initially, tumor recurrence was predicted without imputed pathway scores. Subsequently, the SLCG data were augmented with pathway scores imputed from TCGA. The integrative approach aimed to enhance relapse risk prediction performance.

RESULTS

The integrative approach achieved improved relapse risk prediction with the following evaluation metrics: an area under the precision-recall curve (PR-AUC) score of 0.75, an area under the ROC (ROC-AUC) score of 0.80, an F1 score of 0.61, and a Precision of 0.80. The prediction explanation model SHAP (SHapley Additive exPlanations) was employed to explain the machine learning model's predictions.

CONCLUSION

We conclude that our explainable predictive model is a promising tool for oncologists that addresses an unmet clinical need of post-treatment patient stratification based on the relapse risk while also improving the predictive power by incorporating proxy genomic data not available for specific patients.

Cost Analysis of Magnetic Resonance Imaging and Computed Tomography in Cardiology: A Case Study of a University Hospital Complex in the Euro Region

INTRODUCTION

In recent years, several hospitals have incorporated MRI equipment managed directly by their cardiology departments. The aim of our work is to determine the total cost per test of both CT and MRI in the setting of a Cardiology Department of a tertiary hospital.

MATERIALS AND METHODS

The process followed for estimating the costs of CT and MRI tests consists of three phases: (1) Identification of the phases of the testing process; (2) Identification of the resources consumed in carrying out the tests; (3) Quantification and assessment of inputs.

RESULTS

MRI involves higher personnel (EUR 66.03 vs. EUR 49.17) and equipment (EUR 89.98 vs. EUR 33.73) costs, while CT consumes higher expenditures in consumables (EUR 93.28 vs. EUR 22.95) and overheads (EUR 1.64 vs. EUR 1.55). The total cost of performing each test is higher in MRI (EUR 180.60 vs. EUR 177.73).

CONCLUSIONS

We can conclude that the unit cost of each CT and MRI performed in that unit are EUR 177.73 and EUR 180.60, respectively, attributable to consumables in the case of CT and to amortization of equipment and staff time in the case of MRI.

A novel, effective machine learning-based RNA editing profile for predicting the prognosis of lower-grade gliomas

Patients with low-grade glioma (LGG) may survive for long time periods, but their tumors often progress to higher-grade lesions. Currently, no cure for LGG is available. A-to-I RNA editing accounts for nearly 90% of all RNA editing events in humans and plays a role in tumorigenesis in various cancers. However, little is known regarding its prognostic role in LGG. On the basis of The Cancer Genome Atlas (TCGA) data, we used LASSO and univariate Cox regression to construct an RNA editing site signature. The results derived from the TCGA dataset were further validated with Gene Expression Omnibus (GEO) and Chinese Glioma Genome Atlas (CGGA) datasets. Five machine learning algorithms (Decision Trees C5.0, XGboost, GBDT, Lightgbm, and Catboost) were used to confirm the prognosis associated with the RNA editing site signature. Finally, we explored immune function, immunotherapy, and potential therapeutic agents in the high- and low-risk groups by using multiple biological prediction websites. A total of 22,739 RNA editing sites were identified, and a signature model consisting of four RNA editing sites (PRKCSH|chr19:11561032, DSEL|chr18:65174489, UGGT1|chr2:128952084, and SOD2|chr6:160101723) was established. Cox regression analysis indicated that the RNA editing signature was an independent prognostic factor, according to the ROC curve (AUC = 0.823), and the nomogram model had good predictive power (C-index = 0.824). In addition, the predictive ability of the RNA editing signature was confirmed with the machine learning model. The sensitivity of PCI-34051 and Elephantin was significantly higher in the high-risk group than the low-risk group, thus potentially providing a marker to predict the effects of lung cancer drug treatment. RNA editing may serve as a novel survival prediction tool, thus offering hope for developing editing-based therapeutic strategies to combat LGG progression. In addition, this tool may help optimize survival risk assessment and individualized care for patients with low-grade gliomas.

Use of machine learning to predict bladder cancer survival outcomes: a systematic literature review

INTRODUCTION

The objective of this systematic review is to summarize the use of machine learning (ML) in predicting overall survival (OS) in patients with bladder cancer.

METHODS

Search terms for bladder cancer, ML algorithms, and mortality were used to identify studies in PubMed and Web of Science as of February 2022. Notable inclusion/exclusion criteria contained the inclusion of studies that utilized patient-level datasets and exclusion of primary gene expression-related dataset studies. Study quality and bias were assessed using the International Journal of Medical Informatics (IJMEDI) checklist.

RESULTS

Of the 14 included studies, the most common algorithms were artificial neural networks (n = 8) and logistic regression (n = 4). Nine articles described missing data handling, with five articles removing patients with missing data entirely. With respect to feature selection, the most common sociodemographic variables were age (n = 9), gender (n = 9), and smoking status (n = 3), with clinical variables most commonly including tumor stage (n = 8), grade (n = 7), and lymph node involvement (n = 6). Most studies (n = 10) were of medium IJMEDI quality, with common areas of improvement being the descriptions of data preparation and deployment.

CONCLUSIONS

ML holds promise for optimizing bladder cancer care through accurate OS predictions, but challenges related to data processing, feature selection, and data source quality must be resolved to develop robust models. While this review is limited by its inability to compare models across studies, this systematic review will inform decision-making by various stakeholders to improve understanding of ML-based OS prediction in bladder cancer and foster interpretability of future models.

Prediction of Cervical Lymph Node Metastasis in Clinically Node-Negative T1 and T2 Papillary Thyroid Carcinoma Using Supervised Machine Learning Approach

Papillary thyroid carcinoma (PTC) is generally considered an indolent cancer. However, patients with cervical lymph node metastasis (LNM) have a higher risk of local recurrence. This study evaluated and compared four machine learning (ML)-based classifiers to predict the presence of cervical LNM in clinically node-negative (cN0) T1 and T2 PTC patients. The algorithm was developed using clinicopathological data from 288 patients who underwent total thyroidectomy and prophylactic central neck dissection, with sentinel lymph node biopsy performed to identify lateral LNM. The final ML classifier was selected based on the highest specificity and the lowest degree of overfitting while maintaining a sensitivity of 95%. Among the models evaluated, the k-Nearest Neighbor (k-NN) classifier was found to be the best fit, with an area under the receiver operating characteristic curve of 0.72, and sensitivity, specificity, positive and negative predictive values, F1 and F2 scores of 98%, 27%, 56%, 93%, 72%, and 85%, respectively. A web application based on a sensitivity-optimized kNN classifier was also created to predict the potential of cervical LNM, allowing users to explore and potentially build upon the model. These findings suggest that ML can improve the prediction of LNM in cN0 T1 and T2 PTC patients, thereby aiding in individual treatment planning.

Evaluating Efficiency of a Provincial Telerehabilitation Service in Improving Access to Care During the COVID-19 Pandemic

Scope

Early in the COVID-19 pandemic, community rehabilitation stakeholders from a provincial health system designed a novel telerehabilitation service. The service provided wayfinding and self-management advice to individuals with musculoskeletal concerns, neurological conditions, or post-COVID-19 recovery needs. This study evaluated the efficiency of the service in improving access to care.

Methodology

We used multiple methods including secondary data analyses of call metrics, narrative analyses of clinical notes using artificial intelligence (AI) and machine learning (ML), and qualitative interviews.

Conclusions

Interviews revealed that the telerehabilitation service had the potential to positively impact access to rehabilitation during the COVID-19 pandemic, for individuals living rurally, and for individuals on wait lists. Call metric analyses revealed that efficiency may be enhanced if call handling time was reduced. AI/ML analyses found that pain was the most frequently-mentioned keyword in clinical notes, suggesting an area for additional telerehabilitation resources to ensure efficiency.

Artificial Intelligence in Musculoskeletal Conditions

Artificial intelligence (AI) refers to computer capabilities that resemble human intelligence. AI implies the ability to learn and perform tasks that have not been specifically programmed. Moreover, it is an iterative process involving the ability of computerized systems to capture information, transform it into knowledge, and process it to produce adaptive changes in the environment. A large labeled database is needed to train the AI system and generate a robust algorithm. Otherwise, the algorithm cannot be applied in a generalized way. AI can facilitate the interpretation and acquisition of radiological images. In addition, it can facilitate the detection of trauma injuries and assist in orthopedic and rehabilitative processes. The applications of AI in musculoskeletal conditions are promising and are likely to have a significant impact on the future management of these patients.

Gynecological cancer prognosis using machine learning techniques: A systematic review of last three decades (1990–2022)

OBJECTIVE

Many Computer Aided Prognostic (CAP) systems based on machine learning techniques have been proposed in the field of oncology. The objective of this systematic review was to assess and critically appraise the methodologies and approaches used in predicting the prognosis of gynecological cancers using CAPs.

METHODS

Electronic databases were used to systematically search for studies utilizing machine learning methods in gynecological cancers. Study risk of bias (ROB) and applicability were assessed using the PROBAST tool. 139 studies met the inclusion criteria, of which 71 predicted outcomes for ovarian cancer patients, 41 predicted outcomes for cervical cancer patients, 28 predicted outcomes for uterine cancer patients, and 2 predicted outcomes for gynecological malignancies broadly.

RESULTS

Random forest (22.30 %) and support vector machine (21.58 %) classifiers were used most commonly. Use of clinicopathological, genomic and radiomic data as predictors was observed in 48.20 %, 51.08 % and 17.27 % of studies, respectively, with some studies using multiple modalities. 21.58 % of studies were externally validated. Twenty-three individual studies compared ML and non-ML methods. Study quality was highly variable and methodologies, statistical reporting and outcome measures were inconsistent, preventing generalized commentary or meta-analysis of performance outcomes.

CONCLUSION

There is significant variability in model development when prognosticating gynecological malignancies with respect to variable selection, machine learning (ML) methods and endpoint selection. This heterogeneity prevents meta-analysis and conclusions regarding the superiority of ML methods. Furthermore, PROBAST-mediated ROB and applicability analysis demonstrates concern for the translatability of existing models. This review identifies ways that this can be improved upon in future works to develop robust, clinically translatable models within this promising field.

Using Machine Learning to make nanomaterials sustainable

Sustainable development is a key challenge for contemporary human societies; failure to achieve sustainability could threaten human survival. In this review article, we illustrate how Machine Learning (ML) could support more sustainable development, covering the basics of data gathering through each step of the Environmental Risk Assessment (ERA). The literature provides several examples showing how ML can be employed in most steps of a typical ERA.A key observation is that there are currently no clear guidance for using such autonomous technologies in ERAs or which standards/checks are required. Steering thus seems to be the most important task for supporting the use of ML in the ERA of nano- and smart-materials. Resources should be devoted to developing a strategy for implementing ML in ERA with a strong emphasis on data foundations, methodologies, and the related sensitivities/uncertainties. We should recognise historical errors and biases (e.g., in data) to avoid embedding them during ML programming.

Machine Learning in Colorectal Cancer Risk Prediction from Routinely Collected Data: A Review

The inclusion of machine-learning-derived models in systematic reviews of risk prediction models for colorectal cancer is rare. Whilst such reviews have highlighted methodological issues and limited performance of the models included, it is unclear why machine-learning-derived models are absent and whether such models suffer similar methodological problems. This scoping review aims to identify machine-learning models, assess their methodology, and compare their performance with that found in previous reviews. A literature search of four databases was performed for colorectal cancer prediction and prognosis model publications that included at least one machine-learning model. A total of 14 publications were identified for inclusion in the scoping review. Data was extracted using an adapted CHARM checklist against which the models were benchmarked. The review found similar methodological problems with machine-learning models to that observed in systematic reviews for non-machine-learning models, although model performance was better. The inclusion of machine-learning models in systematic reviews is required, as they offer improved performance despite similar methodological omissions; however, to achieve this the methodological issues that affect many prediction models need to be addressed.

The application of machine learning to imaging in hematological oncology: A scoping review

Background

Here, we conducted a scoping review to (i) establish which machine learning (ML) methods have been applied to hematological malignancy imaging; (ii) establish how ML is being applied to hematological cancer radiology; and (iii) identify addressable research gaps.

Methods

The review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Extension for Scoping Reviews guidelines. The inclusion criteria were (i) pediatric and adult patients with suspected or confirmed hematological malignancy undergoing imaging (population); (ii) any study using ML techniques to derive models using radiological images to apply to the clinical management of these patients (concept); and (iii) original research articles conducted in any setting globally (context). Quality Assessment of Diagnostic Accuracy Studies 2 criteria were used to assess diagnostic and segmentation studies, while the Newcastle-Ottawa scale was used to assess the quality of observational studies.

Results

Of 53 eligible studies, 33 applied diverse ML techniques to diagnose hematological malignancies or to differentiate them from other diseases, especially discriminating gliomas from primary central nervous system lymphomas (n=18); 11 applied ML to segmentation tasks, while 9 applied ML to prognostication or predicting therapeutic responses, especially for diffuse large B-cell lymphoma. All studies reported discrimination statistics, but no study calculated calibration statistics. Every diagnostic/segmentation study had a high risk of bias due to their case-control design; many studies failed to provide adequate details of the reference standard; and only a few studies used independent validation.

Conclusion

To deliver validated ML-based models to radiologists managing hematological malignancies, future studies should (i) adhere to standardized, high-quality reporting guidelines such as the Checklist for Artificial Intelligence in Medical Imaging; (ii) validate models in independent cohorts; (ii) standardize volume segmentation methods for segmentation tasks; (iv) establish comprehensive prospective studies that include different tumor grades, comparisons with radiologists, optimal imaging modalities, sequences, and planes; (v) include side-by-side comparisons of different methods; and (vi) include low- and middle-income countries in multicentric studies to enhance generalizability and reduce inequity.

Mathematical modeling of cancer immunotherapy for personalized clinical translation

Encouraging advances are being made in cancer immunotherapy modeling, especially in the key areas of developing personalized treatment strategies based on individual patient parameters, predicting treatment outcomes and optimizing immunotherapy synergy when used in combination with other treatment approaches. Here we present a focused review of the most recent mathematical modeling work on cancer immunotherapy with a focus on clinical translatability. It can be seen that this field is transitioning from pure basic science to applications that can make impactful differences in patients' lives. We discuss how researchers are integrating experimental and clinical data to fully inform models so that they can be applied for clinical predictions, and present the challenges that remain to be overcome if widespread clinical adaptation is to be realized. Lastly, we discuss the most promising future applications and areas that are expected to be the focus of extensive upcoming modeling studies.

Over-the-Counter Breast Cancer Classification Using Machine Learning and Patient Registration Records

This study aims to determine the feasibility of machine learning (ML) and patient registration record to be utilised to develop an over-the-counter (OTC) screening model for breast cancer risk estimation. Data were retrospectively collected from women who came to the Hospital Universiti Sains Malaysia, Malaysia for breast-related problems. Eight ML models were used: k-nearest neighbour (kNN), elastic-net logistic regression, multivariate adaptive regression splines, artificial neural network, partial least square, random forest, support vector machine (SVM), and extreme gradient boosting. Features utilised for the development of the screening models were limited to information in the patient registration form. The final model was evaluated in terms of performance across a mammographic density. Additionally, the feature importance of the final model was assessed using the model agnostic approach. kNN had the highest Youden J index, precision, and PR-AUC, while SVM had the highest F2 score. The kNN model was selected as the final model. The model had a balanced performance in terms of sensitivity, specificity, and PR-AUC across the mammographic density groups. The most important feature was the age at examination. In conclusion, this study showed that ML and patient registration information are feasible to be used as the OTC screening model for breast cancer.

Oxidative stress-related patterns determination for establishment of prognostic models, and characteristics of tumor microenvironment infiltration

Oxidative stress-mediated excessive accumulation of ROS in the body destroys cell homeostasis and participates in various diseases. However, the relationship between oxidative stress-related genes (ORGs) and tumor microenvironment (TME) in gastric cancer remains poorly understood. For improving the treatment strategy of GC, it is necessary to explore the relationship among them. We describe the changes of ORGs in 732 gastric cancer samples from two data sets. The two different molecular subtypes revealed that the changes of ORGs were associated with clinical features, prognosis, and TME. Subsequently, the OE_score was related to RFS, as confirmed by the correlation between OE_score and TME, TMB, MSI, immunotherapy, stem cell analysis, chemotherapeutic drugs, etc. OE_score can be used as an independent predictive marker for the treatment and prognosis of gastric cancer. Further, a Norman diagram was established to improve clinical practicability. Our research showed a potential role of ORGs in clinical features, prognosis, and tumor microenvironment of gastric cancer. Our research findings broaden the understanding of gastric cancer ORGs as a potential target for individualized treatment of gastric cancer and a new direction to evaluate the prognosis.

Machine Learning for Risk Prediction of Oesophago-Gastric Cancer in Primary Care: Comparison with Existing Risk-Assessment Tools

Simple Summary

Oesophago-gastric cancer is one of the commonest cancers worldwide, yet it can be particularly difficult to diagnose given that initial symptoms are often non-specific and routine screening is not available. Cancer risk-assessment tools, which calculate cancer risk based on symptoms and other risk factors present in the primary care record, can aid decisions on referrals for cancer investigations, facilitating earlier diagnosis. Diagnosing common cancers earlier could help improve survival rates. Using UK primary care electronic health record data, we compared five different machine learning techniques for probabilistic classification of cancer patients against a current widely used UK primary care cancer risk-assessment tool. The machine learning algorithms outperformed the current risk-assessment tool, with a higher overall accuracy and an ability to reasonably identify 11–25% more cancer patients. We conclude that machine-learning-based risk-assessment tools could help better identify suitable patients for further investigation and support earlier diagnosis.

Abstract

Oesophago-gastric cancer is difficult to diagnose in the early stages given its typical non-specific initial manifestation. We hypothesise that machine learning can improve upon the diagnostic performance of current primary care risk-assessment tools by using advanced analytical techniques to exploit the wealth of evidence available in the electronic health record. We used a primary care electronic health record dataset derived from the UK General Practice Research Database (7471 cases; 32,877 controls) and developed five probabilistic machine learning classifiers: Support Vector Machine, Random Forest, Logistic Regression, Naïve Bayes, and Extreme Gradient Boosted Decision Trees. Features included basic demographics, symptoms, and lab test results. The Logistic Regression, Support Vector Machine, and Extreme Gradient Boosted Decision Tree models achieved the highest performance in terms of accuracy and AUROC (0.89 accuracy, 0.87 AUROC), outperforming a current UK oesophago-gastric cancer risk-assessment tool (ogRAT). Machine learning also identified more cancer patients than the ogRAT: 11.0% more with little to no effect on false positives, or up to 25.0% more with a slight increase in false positives (for Logistic Regression, results threshold-dependent). Feature contribution estimates and individual prediction explanations indicated clinical relevance. We conclude that machine learning could improve primary care cancer risk-assessment tools, potentially helping clinicians to identify additional cancer cases earlier. This could, in turn, improve survival outcomes.

Development of a Prediction Model for COVID-19 Acute Respiratory Distress Syndrome in Patients With Rheumatic Diseases: Results From the Global Rheumatology Alliance Registry

OBJECTIVE

Some patients with rheumatic diseases might be at higher risk for coronavirus disease 2019 (COVID-19) acute respiratory distress syndrome (ARDS). We aimed to develop a prediction model for COVID-19 ARDS in this population and to create a simple risk score calculator for use in clinical settings.

METHODS

Data were derived from the COVID-19 Global Rheumatology Alliance Registry from March 24, 2020, to May 12, 2021. Seven machine learning classifiers were trained on ARDS outcomes using 83 variables obtained at COVID-19 diagnosis. Predictive performance was assessed in a US test set and was validated in patients from four countries with independent registries using area under the curve (AUC), accuracy, sensitivity, and specificity. A simple risk score calculator was developed using a regression model incorporating the most influential predictors from the best performing classifier.

RESULTS

The study included 8633 patients from 74 countries, of whom 523 (6%) had ARDS. Gradient boosting had the highest mean AUC (0.78; 95% confidence interval [CI]: 0.67-0.88) and was considered the top performing classifier. Ten predictors were identified as key risk factors and were included in a regression model. The regression model that predicted ARDS with 71% (95% CI: 61%-83%) sensitivity in the test set, and with sensitivities ranging from 61% to 80% in countries with independent registries, was used to develop the risk score calculator.

CONCLUSION

We were able to predict ARDS with good sensitivity using information readily available at COVID-19 diagnosis. The proposed risk score calculator has the potential to guide risk stratification for treatments, such as monoclonal antibodies, that have potential to reduce COVID-19 disease progression.

Machine learning models in the prediction of 1-year mortality in patients with advanced hepatocellular cancer on immunotherapy: a proof-of-concept study

INTRODUCTION

Immunotherapy is a new promising treatment for patients with advanced hepatocellular carcinoma (HCC), but is costly and potentially associated with considerable side effects. This study aimed to evaluate the role of machine learning (ML) models in predicting the 1-year cancer-related mortality in advanced HCC patients treated with immunotherapy.

METHOD

395 HCC patients who had received immunotherapy (including nivolumab, pembrolizumab or ipilimumab) between 2014 and 2019 in Hong Kong were included. The whole data sets were randomly divided into training (n = 316) and internal validation (n = 79) set. The data set, including 47 clinical variables, was used to construct six different ML models in predicting the risk of 1-year mortality. The performances of ML models were measured by the area under receiver operating characteristic curve (AUC) and their performances were compared with C-Reactive protein and Alpha Fetoprotein in ImmunoTherapY score (CRAFITY) and albumin-bilirubin (ALBI) score. The ML models were further validated with an external cohort between 2020 and 2021.

RESULTS

The 1-year cancer-related mortality was 51.1%. Of the six ML models, the random forest (RF) has the highest AUC of 0.92 (95% CI 0.87-0.98), which was better than logistic regression (0.82, p = 0.01) as well as the CRAFITY (0.68, p < 0.01) and ALBI score (0.84, p = 0.04). RF had the lowest false positive (2.0%) and false negative rate (5.2%), and performed better than CRAFITY score in the external validation cohort (0.91 vs 0.66, p < 0.01). High baseline AFP, bilirubin and alkaline phosphatase were three common risk factors identified by all ML models.

CONCLUSION

ML models could predict 1-year cancer-related mortality in HCC patients treated with immunotherapy, which may help to select patients who would benefit from this treatment.

Exploring the Utility of Anonymized EHR Datasets in Machine Learning Experiments in the Context of the MODELHealth Project

The object of this paper was the application of machine learning to a clinical dataset that was anonymized using the Mondrian algorithm. (1) Background: The preservation of patient privacy is a necessity rising from the increasing digitization of health data; however, the effect of data anonymization on the performance of machine learning models remains to be explored. (2) Methods: The original EHR derived dataset was subjected to anonymization by applying the Mondrian algorithm for various k values and quasi identifier (QI) set attributes. The logistic regression, decision trees, k-nearest neighbors, Gaussian naive Bayes and support vector machine models were applied to the different dataset versions. (3) Results: The classifiers demonstrated different degrees of resilience to the anonymization, with the decision tree and the KNN models showing remarkably stable performance, as opposed to the Gaussian naïve Bayes model. The choice of the QI set attributes and the generalized information loss value played a more important role than the size of the QI set or the k value. (4) Conclusions: Data anonymization can reduce the performance of certain machine learning models, although the appropriate selection of classifier and parameter values can mitigate this effect.

Translating Data Analytics Into Improved Spine Surgery Outcomes: A Roadmap for Biomedical Informatics Research in 2021

STUDY DESIGN

Narrative review.

OBJECTIVES

There is growing interest in the use of biomedical informatics and data analytics tools in spine surgery. Yet despite the rapid growth in research on these topics, few analytic tools have been implemented in routine spine practice. The purpose of this review is to provide a health information technology (HIT) roadmap to help translate data assets and analytics tools into measurable advances in spine surgical care.

METHODS

We conducted a narrative review of PubMed and Google Scholar to identify publications discussing data assets, analytical approaches, and implementation strategies relevant to spine surgery practice.

RESULTS

A variety of data assets are available for spine research, ranging from commonly used datasets, such as administrative billing data, to emerging resources, such as mobile health and biobanks. Both regression and machine learning techniques are valuable for analyzing these assets, and researchers should recognize the particular strengths and weaknesses of each approach. Few studies have focused on the implementation of HIT, and a variety of methods exist to help translate analytic tools into clinically useful interventions. Finally, a number of HIT-related challenges must be recognized and addressed, including stakeholder acceptance, regulatory oversight, and ethical considerations.

CONCLUSIONS

Biomedical informatics has the potential to support the development of new HIT that can improve spine surgery quality and outcomes. By understanding the development life-cycle that includes identifying an appropriate data asset, selecting an analytic approach, and leveraging an effective implementation strategy, spine researchers can translate this potential into measurable advances in patient care.

Posterior Urethral Valves Outcomes Prediction (PUVOP): a machine learning tool to predict clinically relevant outcomes in boys with posterior urethral valves

BACKGROUND

Early kidney and anatomic features may be predictive of future progression and need for additional procedures in patients with posterior urethral valve (PUV). The objective of this study was to use machine learning (ML) to predict clinically relevant outcomes in these patients.

METHODS

Patients diagnosed with PUV with kidney function measurements at our institution between 2000 and 2020 were included. Pertinent clinical measures were abstracted, including estimated glomerular filtration rate (eGFR) at each visit, initial vesicoureteral reflux grade, and renal dysplasia at presentation. ML models were developed to predict clinically relevant outcomes: progression in CKD stage, initiation of kidney replacement therapy (KRT), and need for clean-intermittent catheterization (CIC). Model performance was assessed by concordance index (c-index) and the model was externally validated.

RESULTS

A total of 103 patients were included with a median follow-up of 5.7 years. Of these patients, 26 (25%) had CKD progression, 18 (17%) required KRT, and 32 (31%) were prescribed CIC. Additionally, 22 patients were included for external validation. The ML model predicted CKD progression (c-index = 0.77; external C-index = 0.78), KRT (c-index = 0.95; external C-index = 0.89) and indicated CIC (c-index = 0.70; external C-index = 0.64), and all performed better than Cox proportional-hazards regression. The models have been packaged into a simple easy-to-use tool, available at https://share.streamlit.io/jcckwong/puvop/main/app.py CONCLUSION: ML-based approaches for predicting clinically relevant outcomes in PUV are feasible. Further validation is warranted, but this implementable model can act as a decision-making aid. A higher resolution version of the Graphical abstract is available as Supplementary information.

Role of artificial intelligence in early detection and screening for pancreatic adenocarcinoma

Pancreatic adenocarcinoma remains to be one of the deadliest malignancies in the world despite treatment advancement over the past few decades. Its low survival rates and poor prognosis can be attributed to ambiguity in recommendations for screening and late symptom onset, contributing to its late presentation. In the recent years, artificial intelligence (AI) as emerged as a field to aid in the process of clinical decision making. Considerable efforts have been made in the realm of AI to screen for and predict future development of pancreatic ductal adenocarcinoma. This review discusses the use of AI in early detection and screening for pancreatic adenocarcinoma, and factors which may limit its use in a clinical setting.

Applications of Artificial Intelligence in Screening, Diagnosis, Treatment, and Prognosis of Colorectal Cancer

Colorectal cancer (CRC) is one of the most common cancers worldwide. Accurate early detection and diagnosis, comprehensive assessment of treatment response, and precise prediction of prognosis are essential to improve the patients’ survival rate. In recent years, due to the explosion of clinical and omics data, and groundbreaking research in machine learning, artificial intelligence (AI) has shown a great application potential in clinical field of CRC, providing new auxiliary approaches for clinicians to identify high-risk patients, select precise and personalized treatment plans, as well as to predict prognoses. This review comprehensively analyzes and summarizes the research progress and clinical application value of AI technologies in CRC screening, diagnosis, treatment, and prognosis, demonstrating the current status of the AI in the main clinical stages. The limitations, challenges, and future perspectives in the clinical implementation of AI are also discussed.

Machine Learning-Based Risk Prediction of Critical Care Unit Admission for Advanced Stage High Grade Serous Ovarian Cancer Patients Undergoing Cytoreductive Surgery: The Leeds-Natal Score

Achieving complete surgical cytoreduction in advanced stage high grade serous ovarian cancer (HGSOC) patients warrants an availability of Critical Care Unit (CCU) beds. Machine Learning (ML) could be helpful in monitoring CCU admissions to improve standards of care. We aimed to improve the accuracy of predicting CCU admission in HGSOC patients by ML algorithms and developed an ML-based predictive score. A cohort of 291 advanced stage HGSOC patients with fully curated data was selected. Several linear and non-linear distances, and quadratic discriminant ML methods, were employed to derive prediction information for CCU admission. When all the variables were included in the model, the prediction accuracies were higher for linear discriminant (0.90) and quadratic discriminant (0.93) methods compared with conventional logistic regression (0.84). Feature selection identified pre-treatment albumin, surgical complexity score, estimated blood loss, operative time, and bowel resection with stoma as the most significant prediction features. The real-time prediction accuracy of the Graphical User Interface CCU calculator reached 95%. Limited, potentially modifiable, mostly intra-operative factors contributing to CCU admission were identified and suggest areas for targeted interventions. The accurate quantification of CCU admission patterns is critical information when counseling patients about peri-operative risks related to their cytoreductive surgery.

Dengue disease dynamics are modulated by the combined influences of precipitation and landscape: A machine learning approach

BACKGROUND

Dengue is an endemic vector-borne disease influenced by environmental factors such as landscape and climate. Previous studies separately assessed the effects of landscape and climate factors on mosquito occurrence and dengue incidence. However, both factors concurrently coexist in time and space and can interact, affecting mosquito development and dengue disease transmission. For example, eggs laid in a suitable environment can hatch after being submerged in rain water. It has been difficult for conventional statistical modeling approaches to demonstrate these combined influences due to mathematical constraints.

OBJECTIVES

To investigate the combined influences of landscape and climate factors on mosquito occurrence and dengue incidence.

METHODS

Entomological, epidemiological, and landscape data from the rainy season (July-December) were obtained from respective government agencies in Metropolitan Manila, Philippines, from 2012 to 2014. Temperature, precipitation and vegetation data were obtained through remote sensing. A random forest algorithm was used to select the landscape and climate variables. Afterward, using the identified key variables, a model-based (MOB) recursive partitioning was implemented to test the combined influences of landscape and climate factors on ovitrap index (vector mosquito occurrence) and dengue incidence.

RESULTS

The MOB recursive partitioning for ovitrap index indicated a high sensitivity of vector mosquito occurrence on environmental conditions generated by a combination of high residential density areas with low precipitation. Moreover, the MOB recursive partitioning indicated high sensitivity of dengue incidence to the effects of precipitation in areas with high proportions of residential density and commercial areas.

CONCLUSIONS

Dengue dynamics are not solely influenced by individual effects of either climate or landscape, but rather by their synergistic or combined effects. The presented findings have the potential to target vector surveillance in areas identified as suitable for mosquito occurrence under specific climatic conditions and may be relevant as part of urban planning strategies to control dengue.

Machine Learning in the Differentiation of Soft Tissue Neoplasms: Comparison of Fat-Suppressed T2WI and Apparent Diffusion Coefficient (ADC) Features-Based Models

Machine learning has been widely used in the characterization of tumors recently. This article aims to explore the feasibility of the whole tumor fat-suppressed (FS) T2WI and ADC features-based least absolute shrinkage and selection operator (LASSO)-logistic predictive models in the differentiation of soft tissue neoplasms (STN). The clinical and MR findings of 160 cases with 161 histologically proven STN were reviewed, retrospectively, 75 with diffusion-weighted imaging (DWI with b values of 50, 400, and 800 s/mm²). They were divided into benign and malignant groups and further divided into training (70%) and validation (30%) cohorts. The MR FS T2WI and ADC features-based LASSO-logistic models were built and compared. The AUC of the FS T2WI features-based LASSO-logistic regression model for benign and malignant prediction was 0.65 and 0.75 for the training and validation cohorts. The model's sensitivity, specificity, and accuracy of the validation cohort were 55%, 96%, and 76.6%. While the AUC of the ADC features-based model was 0.932 and 0.955 for the training and validation cohorts. The model's sensitivity, specificity, and accuracy were 83.3%, 100%, and 91.7%. The performances of these models were also validated by decision curve analysis (DCA). The AUC of the whole tumor ADC features-based LASSO-logistic regression predictive model was larger than that of FS T2WI features (p = 0.017). The whole tumor fat-suppressed T2WI and ADC features-based LASSO-logistic predictive models both can serve as useful tools in the differentiation of STN. ADC features-based LASSO-logistic regression predictive model did better than that of FS T2WI features.

Predicting the risk of cancer in adults using supervised machine learning: a scoping review

OBJECTIVES

The purpose of this scoping review is to: (1) identify existing supervised machine learning (ML) approaches on the prediction of cancer in asymptomatic adults; (2) to compare the performance of ML models with each other and (3) to identify potential gaps in research.

DESIGN

Scoping review using the population, concept and context approach.

SEARCH STRATEGY

PubMed search engine was used from inception to 10 November 2020 to identify literature meeting following inclusion criteria: (1) a general adult (≥18 years) population, either sex, asymptomatic (population); (2) any study using ML techniques to derive predictive models for future cancer risk using clinical and/or demographic and/or basic laboratory data (concept) and (3) original research articles conducted in all settings in any region of the world (context).

RESULTS

The search returned 627 unique articles, of which 580 articles were excluded because they did not meet the inclusion criteria, were duplicates or were related to benign neoplasm. Full-text reviews were conducted for 47 articles and a final set of 10 articles were included in this scoping review. These 10 very heterogeneous studies used ML to predict future cancer risk in asymptomatic individuals. All studies reported area under the receiver operating characteristics curve (AUC) values as metrics of model performance, but no study reported measures of model calibration.

CONCLUSIONS

Research gaps that must be addressed in order to deliver validated ML-based models to assist clinical decision-making include: (1) establishing model generalisability through validation in independent cohorts, including those from low-income and middle-income countries; (2) establishing models for all cancer types; (3) thorough comparisons of ML models with best available clinical tools to ensure transparency of their potential clinical utility; (4) reporting of model calibration performance and (5) comparisons of different methods on the same cohort to reveal important information about model generalisability and performance.

Progression-Free Survival Prediction in Patients with Nasopharyngeal Carcinoma after Intensity-Modulated Radiotherapy: Machine Learning vs. Traditional Statistics

Background: The Cox proportional hazards (CPH) model is the most commonly used statistical method for nasopharyngeal carcinoma (NPC) prognostication. Recently, machine learning (ML) models are increasingly adopted for this purpose. However, only a few studies have compared the performances between CPH and ML models. This study aimed at comparing CPH with two state-of-the-art ML algorithms, namely, conditional survival forest (CSF) and DeepSurv for disease progression prediction in NPC. Methods: From January 2010 to March 2013, 412 eligible NPC patients were reviewed. The entire dataset was split into training cohort and testing cohort in a ratio of 90%:10%. Ten features from patient-related, disease-related, and treatment-related data were used to train the models for progression-free survival (PFS) prediction. The model performance was compared using the concordance index (c-index), Brier score, and log-rank test based on the risk stratification results. Results: DeepSurv (c-index = 0.68, Brier score = 0.13, log-rank test p = 0.02) achieved the best performance compared to CSF (c-index = 0.63, Brier score = 0.14, log-rank test p = 0.38) and CPH (c-index = 0.57, Brier score = 0.15, log-rank test p = 0.81). Conclusions: Both CSF and DeepSurv outperformed CPH in our relatively small dataset. ML-based survival prediction may guide physicians in choosing the most suitable treatment strategy for NPC patients.

A comparative study of machine learning techniques for suicide attempts predictive model

Current suicide risk assessments for predicting suicide attempts are time consuming, of low predictive value and have inadequate reliability. This paper aims to develop a predictive model for suicide attempts among patients with depression using machine learning algorithms as well as presents a comparative study on single predictive models with ensemble predictive models for differentiating depressed patients with suicide attempts from non-suicide attempters. We applied and trained eight different machine learning algorithms using a dataset that consists of 75 patients diagnosed with a depressive disorder. A recursive feature elimination was used to reduce the features via three-fold cross validation. An ensemble predictive models outperformed the single predictive models. Voting and bagging revealed the highest accuracy of 92% compared to other machine learning algorithms. Our findings indicate that history of suicide attempt, religion, race, suicide ideation and severity of clinical depression are useful factors for prediction of suicide attempts.

Explainable Artificial Intelligence Reveals Novel Insight into Tumor Microenvironment Conditions Linked with Better Prognosis in Patients with Breast Cancer

We investigated the data-driven relationship between immune cell composition in the tumor microenvironment (TME) and the ≥5-year survival rates of breast cancer patients using explainable artificial intelligence (XAI) models. We acquired TCGA breast invasive carcinoma data from the cbioPortal and retrieved immune cell composition estimates from bulk RNA sequencing data from TIMER2.0 based on EPIC, CIBERSORT, TIMER, and xCell computational methods. Novel insights derived from our XAI model showed that B cells, CD8+ T cells, M0 macrophages, and NK T cells are the most critical TME features for enhanced prognosis of breast cancer patients. Our XAI model also revealed the inflection points of these critical TME features, above or below which ≥5-year survival rates improve. Subsequently, we ascertained the conditional probabilities of ≥5-year survival under specific conditions inferred from the inflection points. In particular, the XAI models revealed that the B cell fraction (relative to all cells in a sample) exceeding 0.025, M0 macrophage fraction (relative to the total immune cell content) below 0.05, and NK T cell and CD8+ T cell fractions (based on cancer type-specific arbitrary units) above 0.075 and 0.25, respectively, in the TME could enhance the ≥5-year survival in breast cancer patients. The findings could lead to accurate clinical predictions and enhanced immunotherapies, and to the design of innovative strategies to reprogram the breast TME.

A Classification Approach for Cancer Survivors from Those Cancer-Free, Based on Health Behaviors: Analysis of the Lifelines Cohort

Simple Summary

Health behaviors affect health status in cancer survivors. We aimed to identify such key health behaviors using nonlinear algorithms and compare their classification performance with logistic regression, for distinguishing cancer survivors from those cancer-free in a population-based cohort. We used health behaviors and socioeconomic factors for analysis. Participants from the Lifelines population-based cohort were binary classified as cancer survivors or cancer-free using nonlinear algorithms or logistic regression. Data were collected for 107,624 cancer-free participants and 2760 cancer survivors. Using all variables, algorithms obtained an area under the receiver operator curve (AUC) of 0.75 ± 0.01. Using only health behaviors, the algorithms differentiated cancer survivors from cancer-free participants with AUCs of 0.62 ± 0.01 and 0.60 ± 0.01, respectively. In the case–control analyses, both algorithms produced AUCs of 0.52 ± 0.01. The main distinctive classifier was age. No key health behaviors were identified by linear and nonlinear algorithms to differentiate cancer survivors from cancer-free participants.

Abstract

Health behaviors affect health status in cancer survivors. We hypothesized that nonlinear algorithms would identify distinct key health behaviors compared to a linear algorithm and better classify cancer survivors. We aimed to use three nonlinear algorithms to identify such key health behaviors and compare their performances with that of a logistic regression for distinguishing cancer survivors from those without cancer in a population-based cohort study. We used six health behaviors and three socioeconomic factors for analysis. Participants from the Lifelines population-based cohort were binary classified into a cancer-survivors group and a cancer-free group using either nonlinear algorithms or logistic regression, and their performances were compared by the area under the curve (AUC). In addition, we performed case–control analyses (matched by age, sex, and education level) to evaluate classification performance only by health behaviors. Data were collected for 107,624 cancer free participants and 2760 cancer survivors. Using all variables resulted an AUC of 0.75 ± 0.01, using only six health behaviors, the logistic regression and nonlinear algorithms differentiated cancer survivors from cancer-free participants with AUCs of 0.62 ± 0.01 and 0.60 ± 0.01, respectively. The main distinctive classifier was age. Though not relevant to classification, the main distinctive health behaviors were body mass index and alcohol consumption. In the case–control analyses, algorithms produced AUCs of 0.52 ± 0.01. No key health behaviors were identified by linear and nonlinear algorithms to differentiate cancer survivors from cancer-free participants in this population-based cohort.

Prediction of Incident Cancers in the Lifelines Population-Based Cohort

Simple Summary

The accurate prediction of incident cancers could be relevant to understanding and reducing cancer incidence. The aim of this study was to develop machine learning (ML) models that could predict an incident diagnosis of cancer. Data were available for 116,188 cancer-free participants and 4232 incident cancer cases. The main outcome was an incident cancer (excluding skin cancer) during follow-up assessment in a population-based cohort. The performance of three ML algorithms was evaluated using supervised binary classification to identify incident cancers among participants. An overall area under the receiver operator curve (AUC) < 0.75 was obtained; the highest AUC was for prostate cancer AUC > 0.80. Linear and non-linear ML algorithms including socioeconomic, lifestyle, and clinical variables produced a moderate predictive performance of incident cancers in the Lifelines cohort.

Abstract

Cancer incidence is rising, and accurate prediction of incident cancers could be relevant to understanding and reducing cancer incidence. The aim of this study was to develop machine learning (ML) models that could predict an incident diagnosis of cancer. Participants without any history of cancer within the Lifelines population-based cohort were followed for a median of 7 years. Data were available for 116,188 cancer-free participants and 4232 incident cancer cases. At baseline, socioeconomic, lifestyle, and clinical variables were assessed. The main outcome was an incident cancer during follow-up (excluding skin cancer), based on linkage with the national pathology registry. The performance of three ML algorithms was evaluated using supervised binary classification to identify incident cancers among participants. Elastic net regularization and Gini index were used for variables selection. An overall area under the receiver operator curve (AUC) <0.75 was obtained, the highest AUC value was for prostate cancer (random forest AUC = 0.82 (95% CI 0.77–0.87), logistic regression AUC = 0.81 (95% CI 0.76–0.86), and support vector machines AUC = 0.83 (95% CI 0.78–0.88), respectively); age was the most important predictor in these models. Linear and non-linear ML algorithms including socioeconomic, lifestyle, and clinical variables produced a moderate predictive performance of incident cancers in the Lifelines cohort.

Predicting breast cancer 5-year survival using machine learning: A systematic review

BACKGROUND

Accurately predicting the survival rate of breast cancer patients is a major issue for cancer researchers. Machine learning (ML) has attracted much attention with the hope that it could provide accurate results, but its modeling methods and prediction performance remain controversial. The aim of this systematic review is to identify and critically appraise current studies regarding the application of ML in predicting the 5-year survival rate of breast cancer.

METHODS

In accordance with the PRISMA guidelines, two researchers independently searched the PubMed (including MEDLINE), Embase, and Web of Science Core databases from inception to November 30, 2020. The search terms included breast neoplasms, survival, machine learning, and specific algorithm names. The included studies related to the use of ML to build a breast cancer survival prediction model and model performance that can be measured with the value of said verification results. The excluded studies in which the modeling process were not explained clearly and had incomplete information. The extracted information included literature information, database information, data preparation and modeling process information, model construction and performance evaluation information, and candidate predictor information.

RESULTS

Thirty-one studies that met the inclusion criteria were included, most of which were published after 2013. The most frequently used ML methods were decision trees (19 studies, 61.3%), artificial neural networks (18 studies, 58.1%), support vector machines (16 studies, 51.6%), and ensemble learning (10 studies, 32.3%). The median sample size was 37256 (range 200 to 659820) patients, and the median predictor was 16 (range 3 to 625). The accuracy of 29 studies ranged from 0.510 to 0.971. The sensitivity of 25 studies ranged from 0.037 to 1. The specificity of 24 studies ranged from 0.008 to 0.993. The AUC of 20 studies ranged from 0.500 to 0.972. The precision of 6 studies ranged from 0.549 to 1. All of the models were internally validated, and only one was externally validated.

CONCLUSIONS

Overall, compared with traditional statistical methods, the performance of ML models does not necessarily show any improvement, and this area of research still faces limitations related to a lack of data preprocessing steps, the excessive differences of sample feature selection, and issues related to validation. Further optimization of the performance of the proposed model is also needed in the future, which requires more standardization and subsequent validation.

Development of Machine Learning Models to Predict Platinum Sensitivity of High-Grade Serous Ovarian Carcinoma

To support the implementation of individualized disease management, we aimed to develop machine learning models predicting platinum sensitivity in patients with high-grade serous ovarian carcinoma (HGSOC). We reviewed the medical records of 1002 eligible patients. Patients' clinicopathologic characteristics, surgical findings, details of chemotherapy, treatment response, and survival outcomes were collected. Using the stepwise selection method, based on the area under the receiver operating characteristic curve (AUC) values, six variables associated with platinum sensitivity were selected: age, initial serum CA-125 levels, neoadjuvant chemotherapy, pelvic lymph node status, involvement of pelvic tissue other than the uterus and tubes, and involvement of the small bowel and mesentery. Based on these variables, predictive models were constructed using four machine learning algorithms, logistic regression (LR), random forest, support vector machine, and deep neural network; the model performance was evaluated with the five-fold cross-validation method. The LR-based model performed best at identifying platinum-resistant cases with an AUC of 0.741. Adding the FIGO stage and residual tumor size after debulking surgery did not improve model performance. Based on the six-variable LR model, we also developed a web-based nomogram. The presented models may be useful in clinical practice and research.

Artificial intelligence in cancer research: learning at different levels of data granularity

From genome-scale experimental studies to imaging data, behavioral footprints, and longitudinal healthcare records, the convergence of big data in cancer research and the advances in Artificial Intelligence (AI) is paving the way to develop a systems view of cancer. Nevertheless, this biomedical area is largely characterized by the co-existence of big data and small data resources, highlighting the need for a deeper investigation about the crosstalk between different levels of data granularity, including varied sample sizes, labels, data types, and other data descriptors. This review introduces the current challenges, limitations, and solutions of AI in the heterogeneous landscape of data granularity in cancer research. Such a variety of cancer molecular and clinical data calls for advancing the interoperability among AI approaches, with particular emphasis on the synergy between discriminative and generative models that we discuss in this work with several examples of techniques and applications.

Prevalence of Missing Data in the National Cancer Database and Association With Overall Survival

IMPORTANCE

Cancer registries are important real-world data sources consisting of data abstraction from the medical record; however, patients with unknown or missing data are underrepresented in studies that use such data sources.

OBJECTIVE

To assess the prevalence of missing data and its association with overall survival among patients with cancer.

DESIGN, SETTING, AND PARTICIPANTS

In this retrospective cohort study, all variables within the National Cancer Database were reviewed for missing or unknown values for patients with the 3 most common cancers in the US who received diagnoses from January 1, 2006, to December 31, 2015. The prevalence of patient records with missing data and the association with overall survival were assessed. Data analysis was performed from February to August 2020.

EXPOSURES

Any missing data field within a patient record among 63 variables of interest from more than 130 total variables in the National Cancer Database.

MAIN OUTCOMES AND MEASURES

Prevalence of missing data in the medical records of patients with cancer and associated 2-year overall survival.

RESULTS

A total of 1 198 749 patients with non-small cell lung cancer (mean [SD] age, 68.5 [10.9] years; 628 811 men [52.5%]), 2 120 775 patients with breast cancer (mean [SD] age, 61.0 [13.3] years; 2 101 758 women [99.1%]), and 1 158 635 patients with prostate cancer (mean [SD] age, 65.2 [9.0] years; 100% men) were included in the analysis. Among those with non-small cell lung cancer, 851 295 patients (71.0%) were missing data for variables of interest; 2-year overall survival was 33.2% for patients with missing data and 51.6% for patients with complete data (P < .001). Among those with breast cancer, 1 161 096 patients (54.7%) were missing data for variables of interest; 2-year overall survival was 93.2% for patients with missing data and 93.9% for patients with complete data (P < .001). Among those with prostate cancer, 460 167 patients (39.7%) were missing data for variables of interest; 2-year overall survival was 91.0% for patients with missing data and 95.6% for patients with complete data (P < .001).

CONCLUSIONS AND RELEVANCE

This study found that within a large cancer registry-based real-world data source, there was a high prevalence of missing data that were unable to be ascertained from the medical record. The prevalence of missing data among patients with cancer was associated with heterogeneous differences in overall survival. Improvements in documentation and data quality are necessary to make optimal use of real-world data for clinical advancements.

AI and Big Data in Healthcare: Towards a More Comprehensive Research Framework for Multimorbidity

Multimorbidity refers to the coexistence of two or more chronic diseases in one person. Therefore, patients with multimorbidity have multiple and special care needs. However, in practice it is difficult to meet these needs because the organizational processes of current healthcare systems tend to be tailored to a single disease. To improve clinical decision making and patient care in multimorbidity, a radical change in the problem-solving approach to medical research and treatment is needed. In addition to the traditional reductionist approach, we propose interactive research supported by artificial intelligence (AI) and advanced big data analytics. Such research approach, when applied to data routinely collected in healthcare settings, provides an integrated platform for research tasks related to multimorbidity. This may include, for example, prediction, correlation, and classification problems based on multiple interaction factors. However, to realize the idea of this paradigm shift in multimorbidity research, the optimization, standardization, and most importantly, the integration of electronic health data into a common national and international research infrastructure is needed. Ultimately, there is a need for the integration and implementation of efficient AI approaches, particularly deep learning, into clinical routine directly within the workflows of the medical professionals.

Introduction to Artificial Intelligence and Machine Learning for Pathology

CONTEXT.—

Recent developments in machine learning have stimulated intense interest in software that may augment or replace human experts. Machine learning may impact pathology practice by offering new capabilities in analysis, interpretation, and outcomes prediction using images and other data. The principles of operation and management of machine learning systems are unfamiliar to pathologists, who anticipate a need for additional education to be effective as expert users and managers of the new tools.

OBJECTIVE.—

To provide a background on machine learning for practicing pathologists, including an overview of algorithms, model development, and performance evaluation; to examine the current status of machine learning in pathology and consider possible roles and requirements for pathologists in local deployment and management of machine learning systems; and to highlight existing challenges and gaps in deployment methodology and regulation.

DATA SOURCES.—

Sources include the biomedical and engineering literature, white papers from professional organizations, government reports, electronic resources, and authors' experience in machine learning. References were chosen when possible for accessibility to practicing pathologists without specialized training in mathematics, statistics, or software development.

CONCLUSIONS.—

Machine learning offers an array of techniques that in recent published results show substantial promise. Data suggest that human experts working with machine learning tools outperform humans or machines separately, but the optimal form for this combination in pathology has not been established. Significant questions related to the generalizability of machine learning systems, local site verification, and performance monitoring remain to be resolved before a consensus on best practices and a regulatory environment can be established.

Prediction models for recurrence and bleeding in patients with venous thromboembolism: A systematic review and critical appraisal

INTRODUCTION

Prediction models for recurrence and bleeding are infrequently used when deciding on anticoagulant treatment duration after venous thromboembolism (VTE) due to concerns about performance and validity. Our aim was to critically appraise these models by systematically summarizing data from derivation and validation studies.

MATERIALS AND METHODS

MEDLINE and CENTRAL were searched until November 15th, 2019. Studies on prediction models for recurrence or bleeding after at least 3 months of anticoagulation in adult patients with VTE were included. The PROBAST, ROBINS-I and RoB2 tools were used to assess risk of bias and applicability.

RESULTS

Selection yielded 18 studies evaluating 8 models for recurrence (7 on development; 9 on validation; 1 update). Generally, models for recurrent VTE appeared to perform poorly to moderately in external validation studies (C-statistics 0.39-0.66, one 0.83). However, impact studies show that HERDOO2 and Vienna prediction model may identify patients with unprovoked VTE at low recurrence risk. Sixteen studies evaluating 14 models for anticoagulation-related bleeding were identified (7 on development; 9 on validation). Although some models seemed promising in development studies, their predictive performance was poor to moderate in external validation (C-statistics 0.52-0.71). All but 3 studies were considered at high risk of bias, mainly due to limitations in the statistical analysis.

CONCLUSIONS

Prognostic models for recurrence and anticoagulation-related bleeding risk often have important methodological limitations and insufficient predictive accuracy. These findings do not support their use in clinical practice to weigh risks of recurrence and bleeding when deciding on continuing anticoagulation after initial treatment of VTE.

Deep Learning applications for COVID-19

This survey explores how Deep Learning has battled the COVID-19 pandemic and provides directions for future research on COVID-19. We cover Deep Learning applications in Natural Language Processing, Computer Vision, Life Sciences, and Epidemiology. We describe how each of these applications vary with the availability of big data and how learning tasks are constructed. We begin by evaluating the current state of Deep Learning and conclude with key limitations of Deep Learning for COVID-19 applications. These limitations include Interpretability, Generalization Metrics, Learning from Limited Labeled Data, and Data Privacy. Natural Language Processing applications include mining COVID-19 research for Information Retrieval and Question Answering, as well as Misinformation Detection, and Public Sentiment Analysis. Computer Vision applications cover Medical Image Analysis, Ambient Intelligence, and Vision-based Robotics. Within Life Sciences, our survey looks at how Deep Learning can be applied to Precision Diagnostics, Protein Structure Prediction, and Drug Repurposing. Deep Learning has additionally been utilized in Spread Forecasting for Epidemiology. Our literature review has found many examples of Deep Learning systems to fight COVID-19. We hope that this survey will help accelerate the use of Deep Learning for COVID-19 research.

Treatment effect prediction with adversarial deep learning using electronic health records

BACKGROUND

Treatment effect prediction (TEP) plays an important role in disease management by ensuring that the expected clinical outcomes are obtained after performing specialized and sophisticated treatments on patients given their personalized clinical status. In recent years, the wide adoption of electronic health records (EHRs) has provided a comprehensive data source for intelligent clinical applications including the TEP investigated in this study.

METHOD

We examined the problem of using a large volume of heterogeneous EHR data to predict treatment effects and developed an adversarial deep treatment effect prediction model to address the problem. Our model employed two auto-encoders for learning the representative and discriminative features of both patient characteristics and treatments from EHR data. The discriminative power of the learned features was further enhanced by decoding the correlational information between the patient characteristics and subsequent treatments by means of a generated adversarial learning strategy. Thereafter, a logistic regression layer was appended on the top of the resulting feature representation layer for TEP.

RESULT

The proposed model was evaluated on two real clinical datasets collected from the cardiology department of a Chinese hospital. In particular, on acute coronary syndrome (ACS) dataset, the proposed adversarial deep treatment effect prediction (ADTEP) (0.662) exhibited 1.4, 2.2, and 6.3% performance gains in terms of the area under the ROC curve (AUC) over deep treatment effect prediction (DTEP) (0.653), logistic regression (LR) (0.648), and support vector machine (SVM) (0.621), respectively. As for heart failure (HF) case study, the proposed ADTEP also outperformed all benchmarks. The experimental results demonstrated that our proposed model achieved competitive performance compared to state-of-the-art models in tackling the TEP problem.

CONCLUSION

In this work, we propose a novel model to address the TEP problem by utilizing a large volume of observational data from EHR. With adversarial learning strategy, our proposed model can further explore the correlational information between patient statuses and treatments to extract more robust and discriminative representation of patient samples from their EHR data. Such representation finally benefits the model on TEP. The experimental results of two case studies demonstrate the superiority of our proposed method compared to state-of-the-art methods.