Clinical decision support systems: need for evidence, need for evaluation

Utilizing deep learning models in an intelligent spiral drawing classification system for Parkinson's disease classification

Introduction

Parkinson's disease (PD) is a neurodegenerative illness that impairs normal human movement. The primary cause of PD is the deficiency of dopamine in the human brain. PD also leads to several other challenges, including insomnia, eating disturbances, excessive sleepiness, fluctuations in blood pressure, sexual dysfunction, and other issues.

Methods

The suggested system is an extremely promising technological strategy that may help medical professionals provide accurate and unbiased disease diagnoses. This is accomplished by utilizing significant and unique traits taken from spiral drawings connected to Parkinson's disease. While PD cannot be cured, early administration of drugs may significantly improve the condition of a patient with PD. An expeditious and accurate clinical classification of PD ensures that efficacious therapeutic interventions can commence promptly, potentially impeding the advancement of the disease and enhancing the quality of life for both patients and their caregivers. Transfer learning models have been applied to diagnose PD by analyzing important and distinctive characteristics extracted from hand-drawn spirals. The studies were carried out in conjunction with a comparison analysis employing 102 spiral drawings. This work enhances current research by analyzing the effectiveness of transfer learning models, including VGG19, InceptionV3, ResNet50v2, and DenseNet169, for identifying PD using hand-drawn spirals.

Results

Transfer machine learning models demonstrate highly encouraging outcomes in providing a precise and reliable classification of PD. Actual results demonstrate that the InceptionV3 model achieved a high accuracy of 89% when learning from spiral drawing images and had a superior receiver operating characteristic (ROC) curve value of 95%.

Discussion

The comparison results suggest that PD identification using these models is currently at the forefront of PD research. The dataset will be enlarged, transfer learning strategies will be investigated, and the system's integration into a comprehensive Parkinson's monitoring and evaluation platform will be looked into as future research areas. The results of this study could lead to a better quality of life for Parkinson's sufferers, individualized treatment, and an early classification.

A systematic review of the applications of Expert Systems (ES) and machine learning (ML) in clinical urology

BACKGROUND

Testing a hypothesis for 'factors-outcome effect' is a common quest, but standard statistical regression analysis tools are rendered ineffective by data contaminated with too many noisy variables. Expert Systems (ES) can provide an alternative methodology in analysing data to identify variables with the highest correlation to the outcome. By applying their effective machine learning (ML) abilities, significant research time and costs can be saved. The study aims to systematically review the applications of ES in urological research and their methodological models for effective multi-variate analysis. Their domains, development and validity will be identified.

METHODS

The PRISMA methodology was applied to formulate an effective method for data gathering and analysis. This study search included seven most relevant information sources: WEB OF SCIENCE, EMBASE, BIOSIS CITATION INDEX, SCOPUS, PUBMED, Google Scholar and MEDLINE. Eligible articles were included if they applied one of the known ML models for a clear urological research question involving multivariate analysis. Only articles with pertinent research methods in ES models were included. The analysed data included the system model, applications, input/output variables, target user, validation, and outcomes. Both ML models and the variable analysis were comparatively reported for each system.

RESULTS

The search identified n = 1087 articles from all databases and n = 712 were eligible for examination against inclusion criteria. A total of 168 systems were finally included and systematically analysed demonstrating a recent increase in uptake of ES in academic urology in particular artificial neural networks with 31 systems. Most of the systems were applied in urological oncology (prostate cancer = 15, bladder cancer = 13) where diagnostic, prognostic and survival predictor markers were investigated. Due to the heterogeneity of models and their statistical tests, a meta-analysis was not feasible.

CONCLUSION

ES utility offers an effective ML potential and their applications in research have demonstrated a valid model for multi-variate analysis. The complexity of their development can challenge their uptake in urological clinics whilst the limitation of the statistical tools in this domain has created a gap for further research studies. Integration of computer scientists in academic units has promoted the use of ES in clinical urological research.

Parkinson’s Disease Detection from Drawing Movements Using Convolutional Neural Networks

Nowadays, an important research effort in healthcare biometrics is finding accurate biomarkers that allow developing medical-decision support tools. These tools help to detect and supervise illnesses like Parkinson’s disease (PD). This paper contributes to this effort by analyzing a convolutional neural network (CNN) for PD detection from drawing movements. This CNN includes two parts: feature extraction (convolutional layers) and classification (fully connected layers). The inputs to the CNN are the module of the Fast Fourier’s transform in the range of frequencies between 0 Hz and 25 Hz. We analyzed the discrimination capability of different directions during drawing movements obtaining the best results for X and Y directions. This analysis was performed using a public dataset: Parkinson Disease Spiral Drawings Using Digitized Graphics Tablet dataset. The best results obtained in this work showed an accuracy of 96.5%, a F1-score of 97.7%, and an area under the curve of 99.2%.

Developing a new intelligent system for the diagnosis of oral medicine with case-based reasoning approach

OBJECTIVE

Since the clinical manifestations of many oral diseases can be quite similar despite the wide variety in etiology and pathology, the differential diagnosis of oral diseases is a complex and challenging process. Intelligent system for differential diagnosis of oral medicine using the artificial intelligence (AI) capabilities helps specialists in achieving differential diagnosis in a wide range of oral diseases.

MATERIALS AND METHODS

First, the essential data elements to design and develop an intelligent system were identified in a cross-sectional descriptive study. The case-based reasoning method was selected to design and implement the system, which consists of three stages: collect the clinical data, construct the cases database, and case-based reasoning cycle. The problem is solved by CBR method in a cycle consisting of four main stages of retrieval, reuse, review, and retention. The evaluation process was conducted in a pilot-based way through the evaluation of the system's performance in the clinical setting and also using the usability assessment questionnaire.

RESULTS

The output of the present project is a web-based intelligent information system, which is developed using the Visual Studio 2015 software. The database of this system is the Microsoft SQL Server version 2012, which has been programmed based on Net framework (version 4.5 or higher) using Visual Basic language. The results of the system evaluation by specialists in clinical settings showed that the system's diagnosis power in different aspects of the disease is influenced by their prevalence and incidence.

CONCLUSIONS

System development using the artificial intelligence capabilities and through the clinical data analysis has potential to help specialists to determine the best diagnostic strategy to achieve a differential diagnosis of a wide range of oral diseases. The results of evaluation present the potential of the system to improve the quality and efficiency of patient care.

Learning health systems need to bridge the ‘two cultures’ of clinical informatics and data science

Background

UK health research policy and plans for population health management are predicated upon transformative knowledge discovery from operational ‘Big Data’. Learning health systems require not only data, but feedback loops of knowledge into changed practice. This depends on knowledge management and application, which in turn depends upon effective system design and implementation. Biomedical informatics is the interdisciplinary field at the intersection of health science, social science and information science and technology that spans this entire scope.

Issues

In the UK, the separate worlds of health data science (bioinformatics, ‘Big Data’) and effective healthcare system design and implementation (clinical informatics, ‘Digital Health’) have operated as ‘two cultures’. Much National Health Service and social care data is of very poor quality. Substantial research funding is wasted on ‘data cleansing’ or by producing very weak evidence. There is not yet a sufficiently powerful professional community or evidence base of best practice to influence the practitioner community or the digital health industry.

Recommendation

The UK needs increased clinical informatics research and education capacity and capability at much greater scale and ambition to be able to meet policy expectations, address the fundamental gaps in the discipline’s evidence base and mitigate the absence of regulation. Independent evaluation of digital health interventions should be the norm, not the exception.

Conclusions

Policy makers and research funders need to acknowledge the existing gap between the ‘two cultures’ and recognise that the full social and economic benefits of digital health and data science can only be realised by accepting the interdisciplinary nature of biomedical informatics and supporting a significant expansion of clinical informatics capacity and capability.

A multicentre integration of a computer-led follow-up of prostate cancer is valid and safe

OBJECTIVE

To test a computer-led follow-up service for prostate cancer in two UK hospitals; the testing aimed to validate the computer expert system in making clinical decisions according to the individual patient's clinical need with a valid model accurately identify patients with disease recurrence or treatment failure based on their blood test and clinical picture.

PATIENTS AND METHODS

A clinical-decision support system (CDSS) was developed from European (European Association of Urology) and national (National Institute for Health and Care Excellence) guidelines along with knowledge acquired from Urologists. This model was then applied in two UK hospitals to review patients after prostate cancer treatment. These patients' data (n = 200) were then reviewed by two independent urology consultants (blinded from the CDSS and the other consultant's rating) and the agreement was calculated by kappa statistics for validation. The second endpoint was to verify the system by estimating the system reliability.

RESULTS

The two individual urology consultants identified 12% and 15% of the patients to have potential disease progression and recommended their referral to urology care. The kappa coefficient for the agreement between the CDSS and the two consultants was 0.81 (P < 0.001) and 0.84 (P < 0.001). The agreement amongst both specialist was also high with k = 0.83 (P < 0.001). The system reliability was estimated on all cases and this demonstrated 100% repeatability of the decisions.

CONCLUSION

A CDSS follow-up is a valid model for providing safe follow-up for prostate cancer.

Opportunities and obstacles using a clinical decision support system for maternal care in Burkina Faso

OBJECTIVE

Maternal and neonatal mortality is high in sub-Saharan Africa. To support Healthcare Workers (HCWs), a computerized decision support system (CDSS) was piloted at six rural maternal care units in Burkina Faso. During the two years of the study period, it was apparent from reports that the CDSS was not used regularly in clinical practice. This study aimed to explore the reasons why HCWs failed to use the CDSS.

METHODS

A workshop, organized as group discussions and a plenary session, was performed with 13 participants to understand their experience with the CDSS and suggest improvements if pertinent. Workshop transcripts were analyzed thematically. Socio-demographic and usage patterns of the CDSS were examined by a questionnaire and analyzed descriptively.

RESULTS

The participants reported that the contextual basic conditions for using the CDSS were not fulfilled. These included unreliable power supply, none user-friendly partograph, the CDSS was not integrated with workflow and staff lacked motivational incentives. Despite these limitations, the HCWs reported learning benefits from guidance and alerts in the CDSS. Using the CDSS enabled them to discover problems earlier as they learned to focus on symptoms to prevent harmful situations.

CONCLUSION

The CDSS was not tailored to the needs and context of the users. The HCWs, defined their needs and suggested how the CDSS should be re-designed. This suggests that the successful and regular usage of any CDSS in rural settings requires the involvement of users throughout the construction and pilot-testing phases and not only during the early prototype design period.

Evaluation of handwriting kinematics and pressure for differential diagnosis of Parkinson's disease

OBJECTIVE

We present the PaHaW Parkinson's disease handwriting database, consisting of handwriting samples from Parkinson's disease (PD) patients and healthy controls. Our goal is to show that kinematic features and pressure features in handwriting can be used for the differential diagnosis of PD.

METHODS AND MATERIAL

The database contains records from 37 PD patients and 38 healthy controls performing eight different handwriting tasks. The tasks include drawing an Archimedean spiral, repetitively writing orthographically simple syllables and words, and writing of a sentence. In addition to the conventional kinematic features related to the dynamics of handwriting, we investigated new pressure features based on the pressure exerted on the writing surface. To discriminate between PD patients and healthy subjects, three different classifiers were compared: K-nearest neighbors (K-NN), ensemble AdaBoost classifier, and support vector machines (SVM).

RESULTS

For predicting PD based on kinematic and pressure features of handwriting, the best performing model was SVM with classification accuracy of Pacc=81.3% (sensitivity Psen=87.4% and specificity of Pspe=80.9%). When evaluated separately, pressure features proved to be relevant for PD diagnosis, yielding Pacc=82.5% compared to Pacc=75.4% using kinematic features.

CONCLUSION

Experimental results showed that an analysis of kinematic and pressure features during handwriting can help assess subtle characteristics of handwriting and discriminate between PD patients and healthy controls.