Machine Learning Algorithms Predict Body Mass Index Using Nonlinear Trimodal Regression Analysis from Computed Tomography Scans

Artificial neural networks for the detection of odontoid fractures using the Konstanz Information Miner Analytics Platform

STUDY DESIGN

An experimental study.

PURPOSE

This study aimed to investigate the potential use of artificial neural networks (ANNs) in the detection of odontoid fractures using the Konstanz Information Miner (KNIME) Analytics Platform that provides a technique for computer-assisted diagnosis using radiographic X-ray imaging.

OVERVIEW OF LITERATURE

In medical image processing, computer-assisted diagnosis with ANNs from radiographic X-ray imaging is becoming increasingly popular. Odontoid fractures are a common fracture of the axis and account for 10%-15% of all cervical fractures. However, a literature review of computer-assisted diagnosis with ANNs has not been made.

METHODS

This study analyzed 432 open-mouth (odontoid) radiographic views of cervical spine X-ray images obtained from dataset repositories, which were used in developing ANN models based on the convolutional neural network theory. All the images contained diagnostic information, including 216 radiographic images of individuals with normal odontoid processes and 216 images of patients with acute odontoid fractures. The model classified each image as either showing an odontoid fracture or not. Specifically, 70% of the images were training datasets used for model training, and 30% were used for testing. KNIME's graphic user interface-based programming enabled class label annotation, data preprocessing, model training, and performance evaluation.

RESULTS

The graphic user interface program by KNIME was used to report all radiographic X-ray imaging features. The ANN model performed 50 epochs of training. The performance indices in detecting odontoid fractures included sensitivity, specificity, F-measure, and prediction error of 100%, 95.4%, 97.77%, and 2.3%, respectively. The model's accuracy accounted for 97% of the area under the receiver operating characteristic curve for the diagnosis of odontoid fractures.

CONCLUSIONS

The ANN models with the KNIME Analytics Platform were successfully used in the computer-assisted diagnosis of odontoid fractures using radiographic X-ray images. This approach can help radiologists in the screening, detection, and diagnosis of acute odontoid fractures.

Testing soft tissue radiodensity parameters interplay with age and self-reported physical activity

Aging well is directly associated to a healthy lifestyle. The focus of this paper is to relate individual wellness with medical image features. Non-linear trimodal regression analysis (NTRA) is a novel method that models the radiodensitometric distributions of x-ray computed tomography (CT) cross-sections. It generates 11 patient-specific parameters that describe the quality and quantity of muscle, fat, and connective tissues. In this research, the relationship of these 11 NTRA parameters with age, physical activity, and lifestyle is investigated in the 3,157 elderly volunteers AGES-I dataset. First, univariate statistical analyses were performed, and subjects were grouped by age and self-reported past (youth–midlife) and present (within 12 months of the survey) physical activity to ascertain which parameters were the most influential. Then, machine learning (ML) analyses were conducted to classify patients using NTRA parameters as input features for three ML algorithms. ML is also used to classify a Lifestyle index using the age groups. This classification analysis yielded robust results with the lifestyle index underlying the relevant differences of the soft tissues between age groups, especially in fat and connective tissue. Univariate statistical models suggested that NTRA parameters may be susceptible to age and differences between past and present physical activity levels. Moreover, for both age and physical activity, lean muscle parameters expressed more significant variation than fat and connective tissues.

Health technology assessment through Six Sigma Methodology to assess cemented and uncemented protheses in total hip arthroplasty

The purpose of this study is to use Health Technology Assessment (HTA) through the Six Sigma (SS) and DMAIC (Define, Measure, Analyse, Improve, Control) problem-solving strategies for comparing cemented and uncemented prostheses in terms of the costs incurred for Total hip arthroplasty (THA) and the length of hospital stay (LOS). Multinomial logistic regression analysis for modelling the data was also performed. Quantitative parameters extracted from gait analysis, electromyography and computed tomography images were used to compare the approaches, but the analysis did not show statistical significance. The variables regarding costs were studied with the Mann-Whitney and Kruskal-Wallis tests. No statistically significant difference between cemented and uncemented prosthesis for the total cost of LOS was found, but the cost of the surgeon had an influence on the overall expenses, affecting the cemented prosthetic approach. The material costs of surgery for the uncemented prosthesis and the cost of theatre of surgery for the cemented prosthesis were the most influential. Multinomial logistic regression identified the Vastus Lateralis variable as statistically significant. The overall accuracy of the model is 93.0%. The use of SS and DMAIC cycle as tools of HTA proved that the cemented and uncemented approaches for THA have similar costs and LOSy.

A health technology assessment between two pharmacological therapies through Six Sigma: the case study of bone cancer

Purpose

Head and neck cancers are multi-factorial diseases that can affect many sides of people's life and are due to a lot of risk factors. According to their characteristics, the treatment can be surgical, use of radiation or chemotherapy. The use of a surgical treatment can lead to surgical infections that are a main theme in medicine. At the University hospital of Naples “Federico II”, two antibiotics were employed to tackle the issue of the infections and they are compared in this paper to find which one implies the lowest length of hospital stay (LOS) and the reduction of infections.

Design/methodology/approach

The Six Sigma methodology and its problem-solving strategy DMAIC (define, measure, analyse, improve, control), already employed in the healthcare sector, were used as a tool of a health technology assessment between two drugs. In this paper the DMAIC roadmap is used to compare the Ceftriaxone (administered to a group of 48 patients) and the association of Cefazolin plus Clindamycin (administered to a group of 45 patients).

Findings

The results show that the LOS of patients treated with Ceftriaxone is lower than those who were treated with the association of Cefazolin plus Clindamycin, the difference is about 41%. Moreover, a lower number of complications and infections was found in patients who received Ceftriaxone. Finally, a greater number of antibiotic shifts was needed by patients treated with Cefazolin plus Clindamycin.

Research limitations/implications

While the paper enhances clearly the advantages for patients' outcomes regarding the LOS and the number of complications, it did not analyse the costs of the two antibiotics.

Practical implications

Employing the Ceftriaxone would allow the Department of Maxillofacial Surgery to obtain lower LOS and a limited number of complications/infections for recovered patients, consequently reducing the hospitalization costs.

Originality/value

There is a double value in this paper: first of all, the comparison between the two antibiotics gives an answer to one of the main issues in medicine that is the reduction of hospital-acquired infections; secondly, the Six Sigma through its DMAIC cycle can be employed also to compare two biomedical technologies as a tool of health technology assessment studies.

Machine learning predictive system based upon radiodensitometric distributions from mid-thigh CT images

The nonlinear trimodal regression analysis (NTRA) method based on radiodensitometric CT images distributions was developed for the quantitative characterization of soft tissue changes according to the lower extremity function of elderly subjects. In this regard, the NTRA method defines 11 subject-specific soft tissue parameters and has illustrated high sensitivity to changes in skeletal muscle form and function. The present work further explores the use of these 11 NTRA parameters in the construction of a machine learning (ML) system to predict body mass index and isometric leg strength using tree-based regression algorithms. Results obtained from these models demonstrate that when using an ML approach, these soft tissue features have a significant predictive value for these physiological parameters. These results further support the use of NTRA-based ML predictive assessment and support the future investigation of other physiological parameters and comorbidities.

Assessing cardiovascular risks from a mid-thigh CT image: a tree-based machine learning approach using radiodensitometric distributions

The nonlinear trimodal regression analysis (NTRA) method based on radiodensitometric CT distributions was recently developed and assessed for the quantification of lower extremity function and nutritional parameters in aging subjects. However, the use of the NTRA method for building predictive models of cardiovascular health was not explored; in this regard, the present study reports the use of NTRA parameters for classifying elderly subjects with coronary heart disease (CHD), cardiovascular disease (CVD), and chronic heart failure (CHF) using multivariate logistic regression and three tree-based machine learning (ML) algorithms. Results from each model were assembled as a typology of four classification metrics: total classification score, classification by tissue type, tissue-based feature importance, and classification by age. The predictive utility of this method was modelled using CHF incidence data. ML models employing the random forests algorithm yielded the highest classification performance for all analyses, and overall classification scores for all three conditions were excellent: CHD (AUCROC: 0.936); CVD (AUCROC: 0.914); CHF (AUCROC: 0.994). Longitudinal assessment for modelling the prediction of CHF incidence was likewise robust (AUCROC: 0.993). The present work introduces a substantial step forward in the construction of non-invasive, standardizable tools for associating adipose, loose connective, and lean tissue changes with cardiovascular health outcomes in elderly individuals.