Evaluation of classification and regression tree (CART) model in weight loss prediction following head and neck cancer radiation therapy

Differentiating Pressure Ulcer Risk Levels through Interpretable Classification Models Based on Readily Measurable Indicators

Pressure ulcers carry a significant risk in clinical practice. This paper proposes a practical and interpretable approach to estimate the risk levels of pressure ulcers using decision tree models. In order to address the common problem of imbalanced learning in nursing classification datasets, various oversampling configurations are analyzed to improve the data quality prior to modeling. The decision trees built are based on three easily identifiable and clinically relevant pressure ulcer risk indicators: mobility, activity, and skin moisture. Additionally, this research introduces a novel tabular visualization method to enhance the usability of the decision trees in clinical practice. Thus, the primary aim of this approach is to provide nursing professionals with valuable insights for assessing the potential risk levels of pressure ulcers, which could support their decision-making and allow, for example, the application of suitable preventive measures tailored to each patient's requirements. The interpretability of the models proposed and their performance, evaluated through stratified cross-validation, make them a helpful tool for nursing care in estimating the pressure ulcer risk level.

Artificial intelligence to predict outcomes of head and neck radiotherapy

Head and neck radiotherapy induces important toxicity, and its efficacy and tolerance vary widely across patients. Advancements in radiotherapy delivery techniques, along with the increased quality and frequency of image guidance, offer a unique opportunity to individualize radiotherapy based on imaging biomarkers, with the aim of improving radiation efficacy while reducing its toxicity. Various artificial intelligence models integrating clinical data and radiomics have shown encouraging results for toxicity and cancer control outcomes prediction in head and neck cancer radiotherapy. Clinical implementation of these models could lead to individualized risk-based therapeutic decision making, but the reliability of the current studies is limited. Understanding, validating and expanding these models to larger multi-institutional data sets and testing them in the context of clinical trials is needed to ensure safe clinical implementation. This review summarizes the current state of the art of machine learning models for prediction of head and neck cancer radiotherapy outcomes.

Enteral nutrition support and treatment toxicities in patients with head and neck cancer receiving definitive or adjuvant helical intensity-modulated radiotherapy with concurrent chemotherapy

BACKGROUND

Enteral nutrition (EN) is often required in patients with head and neck cancer (HNSCC); however, initiation criteria is limited or inconsistent. This study aimed to describe the relationship of treatment toxicities and requirement for EN and investigate toxicity and baseline characteristics association with EN duration.

METHODS

Acute toxicities and baseline characteristics were collected from patients with HNSCC (n = 110) undergoing H-IMRT. Percentage EN contributing to estimated requirements and EN duration were measured.

RESULTS

The threshold for patients needing ≥50% of estimated requirements via EN increased from week 3 to 4 for grade ≥2 oral/pharyngeal mucositis, dysgeusia, thick saliva and nausea, and for grade 3 dysphagia. Patients with grade 2-3 dysphagia had a reduced risk of ceasing EN compared to those with grade 0-1 dysphagia.

CONCLUSIONS

Using acute toxicities in clinical practice may be a useful tool to inform prompt initiation of EN prior to decline in nutritional status and anticipate EN duration.

CART model to classify the drought status of diverse tomato genotypes by VPD, air temperature, and leaf-air temperature difference

Regular water management is crucial for the cultivation of tomato (Solanum lycopersicum L.). Inadequate irrigation leads to water stress and a reduction in tomato yield and quality. Therefore, it is important to develop an efficient classification method of the drought status of tomato for the timely application of irrigation. In this study, a simple classification and regression tree (CART) model that includes air temperature, vapor pressure deficit, and leaf-air temperature difference was established to classify the drought status of three tomato genotypes (i.e., cherry type 'Tainan ASVEG No. 19', large fruits breeding line '108290', and wild accession 'LA2093'). The results indicate that the proposed CART model exhibited a higher predictive sensitivity, specificity, geometric mean, and accuracy performance compared to the logistic model. In addition, the CART model was applicable not only to three tomato genotypes but across vegetative and reproductive stages. Furthermore, while the drought status was divided into low, medium, and high, the CART model provided a higher predictive performance than that of the logistic model. The results suggest that the drought status of tomato can be accurately classified by the proposed CART model. These results will provide a useful tool of the regular water management for tomato cultivation.

2011-2020 trends of urban and regional ammonia in and around Barcelona, NE Spain

It is well established that in environments where NH₃ abundance is limiting in secondary PM_2.5 generation, a reduction of NH₃ emissions can result in an important contribution to air quality control. However, as deduced from open data published by the European Environmental Agency, the availability of measurements of NH₃ concentrations is very scarce, with very few countries in Europe reporting data consistently for extensive periods, this being especially true for urban background sites. In this framework, simultaneous multi-site measurements were carried out in NE (Northeast) Spain from 2011 to 2020, using diffusion tubes. The highest NH₃ concentrations were recorded at the traffic site (5.3 μgm^-3 on average), followed by those measured at the urban background site (2.1 μgm^-3). Mean concentrations at the mountain site were 1.6 μgm^-3, while the lowest concentrations were recorded at the regional site (0.9 μgm^-3). This comparison highlights traffic emissions as an important source of NH₃. A statistically significant time trend of this pollutant was observed at the urban background site, increasing by 9.4% per year. A season-separated analysis also revealed a significant increasing trend at the mountain site during summer periods, probably related with increasing emissions from agricultural/livestock activities. These increases in NH₃ concentrations were hypothesized to be responsible for the lack of a decreasing trend of NO₃^- concentrations at the monitoring sites, in spite of a markedly reduction of NO₂ during the period, especially at the urban background. Thus, this would in turn affect the effectiveness of current action plans to abate fine aerosols, largely made up of secondary compounds. Actions to reduce NH₃ concentrations at urban backgrounds are challenging though, as predicting NH₃ is subjected to a high uncertainty and complexity due to its dependence on a variety of factors. This complexity was clearly indicated by the application of a decision tree algorithm to find the parameters better predicting NH₃ at the urban background under study. O₃, NO, NO₂, CO, SO₂ and OM + EC concentrations, together with meteorological indicators, were used as independent variables, obtaining no combination of parameters evidently able to predict significant differences in NH₃ concentrations, with a coefficient of determination between real and predicted measurements lower than 0.50. This emphasizes the need for highly temporally and spatially resolved NH₃ measurements for an accurate design of abatement actions.

Machine Learning for Head and Neck Cancer: A Safe Bet?-A Clinically Oriented Systematic Review for the Radiation Oncologist

BACKGROUND AND PURPOSE

Machine learning (ML) is emerging as a feasible approach to optimize patients' care path in Radiation Oncology. Applications include autosegmentation, treatment planning optimization, and prediction of oncological and toxicity outcomes. The purpose of this clinically oriented systematic review is to illustrate the potential and limitations of the most commonly used ML models in solving everyday clinical issues in head and neck cancer (HNC) radiotherapy (RT).

MATERIALS AND METHODS

Electronic databases were screened up to May 2021. Studies dealing with ML and radiomics were considered eligible. The quality of the included studies was rated by an adapted version of the qualitative checklist originally developed by Luo et al. All statistical analyses were performed using R version 3.6.1.

RESULTS

Forty-eight studies (21 on autosegmentation, four on treatment planning, 12 on oncological outcome prediction, 10 on toxicity prediction, and one on determinants of postoperative RT) were included in the analysis. The most common imaging modality was computed tomography (CT) (40%) followed by magnetic resonance (MR) (10%). Quantitative image features were considered in nine studies (19%). No significant differences were identified in global and methodological scores when works were stratified per their task (i.e., autosegmentation).

DISCUSSION AND CONCLUSION

The range of possible applications of ML in the field of HN Radiation Oncology is wide, albeit this area of research is relatively young. Overall, if not safe yet, ML is most probably a bet worth making.

Radiation-Induced Skin Dermatitis: Treatment With CamWell® Herb to Soothe® Cream in Patients With Head and Neck Cancer Receiving Radiation Therapy

BACKGROUND

Radiation-induced skin dermatitis (RISD) is a common outcome experienced by adult patients with head and neck cancer (HNC) who have undergone radiation therapy. There is no standardized recommended agent for the prevention or management of RISD.

OBJECTIVES

The primary objective of this study was to retrospectively evaluate for effectiveness of a botanical topical agent, CamWell® Herb to Soothe® cream, on RISD.

METHODS

112 patients with HNC undergoing radiation therapy self-reported their RISD topical skin care agent during treatment as standard of care, CamWell used prophylactically, or CamWell use started after the first week of treatment. The primary endpoint was impact of RISD on the patient, as measured by mean Skindex-16 score throughout treatment. Measures were completed weekly.

FINDINGS

The mean Skindex score was statistically significantly lower for the prophylactic group than for the standard-of-care group. CamWell may have played a role in managing RISD when compared to standard-of-care agents.

Co-clinical FDG-PET radiomic signature in predicting response to neoadjuvant chemotherapy in triple-negative breast cancer

Purpose

We sought to exploit the heterogeneity afforded by patient-derived tumor xenografts (PDX) to first, optimize and identify robust radiomic features to predict response to therapy in subtype-matched triple negative breast cancer (TNBC) PDX, and second, to implement PDX-optimized image features in a TNBC co-clinical study to predict response to therapy using machine learning (ML) algorithms.

Methods

TNBC patients and subtype-matched PDX were recruited into a co-clinical FDG-PET imaging trial to predict response to therapy. One hundred thirty-one imaging features were extracted from PDX and human-segmented tumors. Robust image features were identified based on reproducibility, cross-correlation, and volume independence. A rank importance of predictors using ReliefF was used to identify predictive radiomic features in the preclinical PDX trial in conjunction with ML algorithms: classification and regression tree (CART), Naïve Bayes (NB), and support vector machines (SVM). The top four PDX-optimized image features, defined as radiomic signatures (RadSig), from each task were then used to predict or assess response to therapy. Performance of RadSig in predicting/assessing response was compared to SUV_mean, SUV_max, and lean body mass-normalized SUL_peak measures.

Results

Sixty-four out of 131 preclinical imaging features were identified as robust. NB-RadSig performed highest in predicting and assessing response to therapy in the preclinical PDX trial. In the clinical study, the performance of SVM-RadSig and NB-RadSig to predict and assess response was practically identical and superior to SUV_mean, SUV_max, and SUL_peak measures.

Conclusions

We optimized robust FDG-PET radiomic signatures (RadSig) to predict and assess response to therapy in the context of a co-clinical imaging trial.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00259-021-05489-8.

Models for Studying the Distribution of Ticks and Tick-Borne Diseases in Animals: A Systematic Review and a Meta-Analysis with a Focus on Africa

Ticks and tick-borne diseases (TTBD) are constraints to the development of livestock and induce potential human health problems. The worldwide distribution of ticks is not homogenous. Some places are ecologically suitable for ticks but they are not introduced in these areas yet. The absence or low density of hosts is a factor affecting the dissemination of the parasite. To understand the process of introduction and spread of TTBD in different areas, and forecast their presence, scientists developed different models (e.g., predictive models and explicative models). This study aimed to identify models developed by researchers to analyze the TTBD distribution and to assess the performance of these various models with a meta-analysis. A literature search was implemented with PRISMA protocol in two online databases (Scopus and PubMed). The selected articles were classified according to country, type of models and the objective of the modeling. Sensitivity, specificity and accuracy available data of these models were used to evaluate their performance using a meta-analysis. One hundred studies were identified in which seven tick genera were modeled, with Ixodes the most frequently modeled. Additionally, 13 genera of tick-borne pathogens were also modeled, with Borrelia the most frequently modeled. Twenty-three different models were identified and the most frequently used are the generalized linear model representing 26.67% and the maximum entropy model representing 24.17%. A focus on TTBD modeling in Africa showed that, respectively, genus Rhipicephalus and Theileria parva were the most modeled. A meta-analysis on the quality of 20 models revealed that maximum entropy, linear discriminant analysis, and the ecological niche factor analysis models had, respectively, the highest sensitivity, specificity, and area under the curve effect size among all the selected models. Modeling TTBD is highly relevant for predicting their distribution and preventing their adverse effect on animal and human health and the economy. Related results of such analyses are useful to build prevention and/or control programs by veterinary and public health authorities.

Development and Validation of the Predictive Model for Esophageal Squamous Cell Carcinoma Differentiation Degree

The diagnosis of the degree of differentiation of tumor cells can help physicians to make timely detection and take appropriate treatment for the patient's condition. In this study, the original dataset is clustered into two independent types by the Kohonen clustering algorithm. One type is used as the development sets to find correlation indicators and establish predictive models of differentiation, while the other type is used as the validation sets to test the correlation indicators and models. In the development sets, thirteen indicators significantly associated with the degree of differentiation of esophageal squamous cell carcinoma are found by the Kohonen clustering algorithm. Thirteen relevant indicators are used as input features and the degree of tumor differentiations is used as output. Ten classification algorithms are used to predict the differentiation of esophageal squamous cell carcinoma. Artificial bee colony-support vector machine (ABC-SVM) predicts better than the other nine algorithms, with an average accuracy of 81.5% for the 10-fold cross-validation. Based on logistic regression and ReliefF algorithm, five models with the greater merit for the degree of differentiation are found in the development sets. The AUC values of the five models are 0.672, 0.628, 0.630, 0.628, and 0.608 (P < 0.05). The AUC values of the five models in the validation sets are 0.753, 0.728, 0.744, 0.776, and 0.868 (P < 0.0001). The predicted values of the five models are constructed as the input features of ABC-SVM. The accuracy of the 10-fold cross-validation reached 82.0 and 86.5% in the development sets and the validation sets, respectively.

Patient Risk Factor Profiles Associated With the Timing of Goals-of-Care Consultation Before Death: A Classification and Regression Tree Analysis

Background:

Early palliative care consultation (“PCC”) to discuss goals-of-care benefits seriously ill patients. Risk factor profiles associated with the timing of conversations in hospitals, where late conversations most likely occur, are needed.

Objective:

To identify risk factor patient profiles associated with PCC timing before death.

Methods:

Secondary analysis of an observational study was conducted at an urban, academic medical center. Patients aged 18 years and older admitted to the medical center, who had PCC, and died July 1, 2014 to October 31, 2016, were included. Patients admitted for childbirth or rehabilitationand patients whose date of death was unknown were excluded. Classification and Regression Tree modeling was employed using demographic and clinical variables.

Results:

Of 1141 patients, 54% had PCC “close to death” (0-14 days before death); 26% had PCC 15 to 60 days before death; 21% had PCC >60 days before death (median 13 days before death). Variables associated with receiving PCC close to death included being Hispanic or “Other” race/ethnicity intensive care patients with extreme illness severity (85%), with age <46 or >75 increasing this probability (98%). Intensive care patients with extreme illness severity were also likely to receive PCC close to death (64%) as were 50% of intensive care patients with less than extreme illness severity.

Conclusions:

A majority of patients received PCC close to death. A complex set of variable interactions were associated with PCC timing. A systematic process for engaging patients with PCC earlier in the care continuum, and in intensive care regardless of illness severity, is needed.

Predicting acute radiation induced xerostomia in head and neck Cancer using MR and CT Radiomics of parotid and submandibular glands

Purpose

To analyze baseline CT/MR-based image features of salivary glands to predict radiation-induced xerostomia 3-months after head-and-neck cancer (HNC) radiotherapy.

Methods

A retrospective analysis was performed on 266 HNC patients who were treated using radiotherapy at our institution between 2009 and 2018. CT and T1 post-contrast MR images along with NCI-CTCAE xerostomia grade (3-month follow-up) were prospectively collected at our institution. CT and MR images were registered on which parotid/submandibular glands were contoured. Image features were extracted for ipsilateral/contralateral parotid and submandibular glands relative to the location of the primary tumor. Dose-volume-histogram (DVH) parameters were also acquired. Features were pre-selected based on Spearman correlation before modelling by examining the correlation with xerostomia (p < 0.05). A shrinkage regression analysis of the pre-selected features was performed using LASSO. The internal validity of the variable selection was estimated by repeating the entire variable selection procedure using a leave-one-out-cross-validation. The most frequently selected variables were considered in the final model. A generalized linear regression with repeated ten-fold cross-validation was developed to predict radiation-induced xerostomia at 3-months after radiotherapy. This model was tested in an independent dataset (n = 50) of patients who were treated at the same institution in 2017–2018. We compared the prediction performances under eight conditions (DVH-only, CT-only, MR-only, CT + MR, DVH + CT, DVH + CT + MR, Clinical+CT + MR, and Clinical+DVH + CT + MR) using the area under the receiver operating characteristic curve (ROC-AUC).

Results

Among extracted features, 7 CT, 5 MR, and 2 DVH features were selected. The internal cohort (n = 216) ROC-AUC values for DVH, CT, MR, and Clinical+DVH + CT + MR features were 0.73 ± 0.01, 0.69 ± 0.01, 0.70 ± 0.01, and 0.79 ± 0.01, respectively. The validation cohort (n = 50) ROC-AUC values for DVH, CT, MR, and Clinical+DVH + CT + MR features were 0.63, 0.57, 0.66, and 0.68, respectively. The DVH-ROC was not significantly different than the CT-ROC (p = 0.8) or MR-ROC (p = 0.4). However, the CT + MR-ROC was significantly different than the CT-ROC (p = 0.03), but not the Clinical+DVH + CT + MR model (p = 0.5).

Conclusion

Our results suggest that baseline CT and MR image features may reflect baseline salivary gland function and potential risk for radiation injury. The integration of baseline image features into prediction models has the potential to improve xerostomia risk stratification with the ultimate goal of truly personalized HNC radiotherapy.

Electronic supplementary material

The online version of this article (10.1186/s13014-019-1339-4) contains supplementary material, which is available to authorized users.

Balancing accuracy and interpretability of machine learning approaches for radiation treatment outcomes modeling

Radiation outcomes prediction (ROP) plays an important role in personalized prescription and adaptive radiotherapy. A clinical decision may not only depend on an accurate radiation outcomes’ prediction, but also needs to be made based on an informed understanding of the relationship among patients’ characteristics, radiation response and treatment plans. As more patients’ biophysical information become available, machine learning (ML) techniques will have a great potential for improving ROP. Creating explainable ML methods is an ultimate task for clinical practice but remains a challenging one. Towards complete explainability, the interpretability of ML approaches needs to be first explored. Hence, this review focuses on the application of ML techniques for clinical adoption in radiation oncology by balancing accuracy with interpretability of the predictive model of interest. An ML algorithm can be generally classified into an interpretable (IP) or non-interpretable (NIP) (“black box”) technique. While the former may provide a clearer explanation to aid clinical decision-making, its prediction performance is generally outperformed by the latter. Therefore, great efforts and resources have been dedicated towards balancing the accuracy and the interpretability of ML approaches in ROP, but more still needs to be done. In this review, current progress to increase the accuracy for IP ML approaches is introduced, and major trends to improve the interpretability and alleviate the “black box” stigma of ML in radiation outcomes modeling are summarized. Efforts to integrate IP and NIP ML approaches to produce predictive models with higher accuracy and interpretability for ROP are also discussed.

Needs and Challenges for Radiation Oncology in the Era of Precision Medicine

Modern medicine, including the care of the cancer patient, has significantly advanced, with the evidence-based medicine paradigm serving to guide clinical care decisions. Yet we now also recognize the tremendous heterogeneity not only of disease states but of the patient and his or her environment as it influences treatment outcomes and toxicities. These reasons and many others have led to a reevaluation of the generalizability of randomized trials and growing interest in accounting for this heterogeneity under the rubric of precision medicine as it relates to personalizing clinical care predictions, decisions, and therapy for the disease state. For the cancer patient treated with radiation therapy, characterizing the spatial treatment heterogeneity has been a fundamental tenet of routine clinical care facilitated by established database and imaging platforms. Leveraging these platforms to further characterize and collate all clinically relevant sources of heterogeneity that affect the longitudinal health outcomes of the irradiated cancer patient provides an opportunity to generate a critical informatics infrastructure on which precision radiation therapy may be realized. In doing so, data science-driven insight discoveries, personalized clinical decisions, and the potential to accelerate translational efforts may be realized ideally within a network of institutions with locally developed yet coordinated informatics infrastructures. The path toward realizing these goals has many needs and challenges, which we summarize, with many still to be realized and understood. Early efforts by our group have identified the feasibility of this approach using routine clinical data sets and offer promise that this transformation can be successfully realized in radiation oncology.

Phosphoinositide 3-kinase-delta could be a biomarker for eosinophilic nasal polyps

Nasal polyps (NP) cause diverse clinical symptoms of chronic rhinosinusitis (CRS). Chronic inflammation of sinonasal mucosa is known to be crucial in NP formation. We aimed to define the implications of phosphoinositide 3-kinase (PI3K)-δ in nasal inflammation associated with NP by analyzing NP tissue obtained from CRS patients. Results showed that expression of p110δ, a regulatory subunit of PI3K-δ, in NP tissue was increased compared to control tissue. Increased p110δ expression was closely correlated with more severe CRS features. Interestingly, p110δ expression was increased in eosinophilic NP, which are closely related to more complicated clinical courses of the disease. Furthermore, CRS patients possessing NP with higher p110δ expression displayed more eosinophils in NP tissue and blood, higher levels of IL-5 in NP tissue, and more severe features of the disease. Therefore, PI3K-δ may contribute to the formation of NP, especially eosinophilic NP associated with more severe clinical presentations and radiological features.

The Needs and Benefits of Continuous Model Updates on the Accuracy of RT-Induced Toxicity Prediction Models Within a Learning Health System

PURPOSE

Clinical data collection and development of outcome prediction models by machine learning can form the foundation for a learning health system offering precision radiation therapy. However, changes in clinical practice over time can affect the measures and patient outcomes and, hence, the collected data. We hypothesize that regular prediction model updates and continuous prospective data collection are important to prevent the degradation of a model's predication accuracy.

METHODS AND MATERIALS

Clinical and dosimetric data from head and neck patients receiving intensity modulated radiation therapy from 2008 to 2015 were prospectively collected as a routine clinical workflow and anonymized for this analysis. Prediction models for grade ≥2 xerostomia at 3 to 6 months of follow-up were developed by bivariate logistic regression using the dose-volume histogram of parotid and submandibular glands. A baseline prediction model was developed with a training data set from 2008 to 2009. The selected predictor variables and coefficients were updated by 4 different model updating methods. (A) The prediction model was updated by using only recent 2-year data and applied to patients in the following test year. (B) The model was updated by increasing the training data set yearly. (C) The model was updated by increasing the training data set on the condition that the area under the curve (AUC) of the recent test year was less than 0.6. (D) The model was not updated. The AUC of the test data set was compared among the 4 model updating methods.

RESULTS

Dose to parotid and submandibular glands and grade of xerostomia showed decreasing trends over the years (2008-2015, 297 patients; P < .001). The AUC of predicting grade ≥2 xerostomia for the initial training data set (2008-2009, 41 patients) was 0.6196. The AUC for the test data set (2010-2015, 256 patients) decreased to 0.5284 when the initial model was not updated (D). However, the AUC was significantly improved by model updates (A: 0.6164; B: 0.6084; P < .05). When the model was conditionally updated, the AUC was 0.6072 (C).

CONCLUSIONS

Our preliminary results demonstrate that updating prediction models with prospective data collection is effective for maintaining the performance of xerostomia prediction. This suggests that a machine learning framework can handle the dynamic changes in a radiation oncology clinical practice and may be an important component for the construction of a learning health system.