Deep learning for AI-based diagnosis of skin-related neglected tropical diseases: A pilot study.

BACKGROUND: Deep learning, which is a part of a broader concept of artificial intelligence (AI) and/or machine learning has achieved remarkable success in vision tasks. While there is growing interest in the use of this technology in diagnostic support for skin-related neglected tropical diseases (skin NTDs), there have been limited studies in this area and fewer focused on dark skin. In this study, we aimed to develop deep learning based AI models with clinical images we collected for five skin NTDs, namely, Buruli ulcer, leprosy, mycetoma, scabies, and yaws, to understand how diagnostic accuracy can or cannot be improved using different models and training patterns. METHODOLOGY: This study used photographs collected prospectively in Côte d'Ivoire and Ghana through our ongoing studies with use of digital health tools for clinical data documentation and for teledermatology. Our dataset included a total of 1,709 images from 506 patients. Two convolutional neural networks, ResNet-50 and VGG-16 models were adopted to examine the performance of different deep learning architectures and validate their feasibility in diagnosis of the targeted skin NTDs. PRINCIPAL FINDINGS: The two models were able to correctly predict over 70% of the diagnoses, and there was a consistent performance improvement with more training samples. The ResNet-50 model performed better than the VGG-16 model. A model trained with PCR confirmed cases of Buruli ulcer yielded 1-3% increase in prediction accuracy across all diseases, except, for mycetoma, over a model which training sets included unconfirmed cases. CONCLUSIONS: Our approach was to have the deep learning model distinguish between multiple pathologies simultaneously-which is close to real-world practice. The more images used for training, the more accurate the diagnosis became. The percentages of correct diagnosis increased with PCR-positive cases of Buruli ulcer. This demonstrated that it may be better to input images from the more accurately diagnosed cases in the training models also for achieving better accuracy in the generated AI models. However, the increase was marginal which may be an indication that the accuracy of clinical diagnosis alone is reliable to an extent for Buruli ulcer. Diagnostic tests also have their flaws, and they are not always reliable. One hope for AI is that it will objectively resolve this gap between diagnostic tests and clinical diagnoses with the addition of another tool. While there are still challenges to be overcome, there is a potential for AI to address the unmet needs where access to medical care is limited, like for those affected by skin NTDs.

The application of digital health as a nursing solution for leprosy patients during the COVID-19 pandemic: A systematic review.

Objectives: To analyse the effect of using digital health technology on leprosy control programmes. Method: The systematic review comprised search on PubMed, Scopus, ScienceDirect, SAGE and ProQuest databases for interventional studies published in English language from 2013 to 2021 which used digital health technology for leprosy contact tracing, active leprosy detection, monitoring of multi-drug therapy and treatment management during the corona virus disease-2019 pandemic A standard risk of bias tool was used to evaluate bias in the studies, and the Joanna Briggs Institute protocol was used to assess the quality of the studies analysed. RESULTS: Of the 205 studies initially identified, 15(7.3%) were analysed in detail. Quasi-experimental studies had a low risk of bias compared to the rest. The e-leprosy framework was being used along with applications based on smartphones and artificial intelligence Digital health technology was found to be practical, accessible and effective in leprosy control programmes. CONCLUSIONS: Studies reported favourable findings regarding the use of digital health technology in services related to leprosy patients.

Artificial intelligence: Its role in dermatopathology.

Artificial intelligence (AI), a major frontier in the field of medical research, can potentially lead to a paradigm shift in clinical practice. A type of artificial intelligence system known as convolutional neural network points to the possible utility of deep learning in dermatopathology. Though pathology has been traditionally restricted to microscopes and glass slides, recent advancement in digital pathological imaging has led to a transition making it a potential branch for the implementation of artificial intelligence. The current application of artificial intelligence in dermatopathology is to complement the diagnosis and requires a well-trained dermatopathologist's guidance for better designing and development of deep learning algorithms. Here we review the recent advances of artificial intelligence in dermatopathology, its applications in disease diagnosis and in research, along with its limitations and future potential.

The role of mobile teledermoscopy in skin cancer triage and management during the COVID-19 pandemic.

The unprecedented onset of the COVID-19 crisis poses a significant challenge to all fields of medicine, including dermatology. Since the start of the coronavirus outbreak, a stark decline in new skin cancer diagnoses has been reported by countries worldwide. One of the greatest challenges during the pandemic has been the reduced access to face-to-face dermatologic evaluation and non-urgent procedures, such as biopsies or surgical excisions. Teledermatology is a well-integrated alternative when face-to-face dermatological assistance is not available. Teledermoscopy, an extension of teledermatology, comprises consulting dermoscopic images to improve the remote assessment of pigmented and non-pigmented lesions when direct visualisation of lesions is difficult. One of teledermoscopy's greatest strengths may be its utility as a triage and monitoring tool, which is critical in the early detection of skin cancer, as it can reduce the number of unnecessary referrals, wait times, and the cost of providing and receiving dermatological care. Mobile teledermoscopy may act as a communication tool between medical practitioners and patients. By using their smartphone (mobile phone) patients can monitor a suspicious skin lesion identified by their medical practitioner, or alternatively self-detect concerning lesions and forward valuable dermoscopic images for remote medical evaluation. Several mobile applications that allow users to photograph suspicious lesions with their smartphones and have them evaluated using artificial intelligence technology have recently emerged. With the growing popularity of mobile apps and consumer-involved healthcare, this will likely be a key component of skin cancer screening in the years to come. However, most of these applications apply artificial intelligence technology to assess clinical images rather than dermoscopic images, which may lead to lower diagnostic accuracy. Incorporating the direct-to-consumer mobile dermoscopy model in combination with mole-scanning artificial intelligence as a mobile app may be the future of skin cancer detection.

Utility of a real-time fluorescence imaging device in guiding antibiotic treatment in superficial skin infections.

The prescription of antibiotics empirically without confirmation of an infective etiology is on the rise. Administration of appropriate antibiotics can be guided by real-time fluorescence imaging using a point-of-care device. These composite images show the presence, type and the burden of infection. The time saved by this method over microbiological testing, especially in resource-poor settings, can lead to a paradigm shift in treatment by facilitating prompt and adequate antimicrobial therapy, surgical debridement as well as follow-up. Thumbnail sketches of a series of four cases highlighting different scenarios in which a fluorescent imaging device utilizing artificial intelligence and machine learning was found useful is presented in this report.

Artificial intelligence in dermatology and healthcare: An overview.

Many aspects of our life are affected by technology. One of the most discussed advancements of modern technologies is artificial intelligence. It involves computational methods which in some way mimic the human thought process. Just like other branches, the medical field also has come under the ambit of artificial intelligence. Almost every field in medicine has been touched by its effect in one way or the other. Prominent among them are medical diagnosis, medical statistics, robotics, and human biology. Medical imaging is one of the foremost specialties with artificial intelligence applications, wherein deep learning methods like artificial neural networks are commonly used. artificial intelligence application in dermatology was initially restricted to the analysis of melanoma and pigmentary skin lesions, has now expanded and covers many dermatoses. Though the applications of artificial intelligence are ever increasing, large data requirements, interpretation of data and ethical concerns are some of its limitations in the present day.

Leprosy Screening Based on Artificial Intelligence: Development of a Cross-Platform App.

BACKGROUND: According to the World Health Organization, achieving targets for control of leprosy by 2030 will require disease elimination and interruption of transmission at the national or regional level. India and Brazil have reported the highest leprosy burden in the last few decades, revealing the need for strategies and tools to help health professionals correctly manage and control the disease. OBJECTIVE: The main objective of this study was to develop a cross-platform app for leprosy screening based on artificial intelligence (AI) with the goal of increasing accessibility of an accurate method of classifying leprosy treatment for health professionals, especially for communities further away from major diagnostic centers. Toward this end, we analyzed the quality of leprosy data in Brazil on the National Notifiable Diseases Information System (SINAN). METHODS: Leprosy data were extracted from the SINAN database, carefully cleaned, and used to build AI decision models based on the random forest algorithm to predict operational classification in paucibacillary or multibacillary leprosy. We used Python programming language to extract and clean the data, and R programming language to train and test the AI model via cross-validation. To allow broad access, we deployed the final random forest classification model in a web app via shinyApp using data available from the Brazilian Institute of Geography and Statistics and the Department of Informatics of the Unified Health System. RESULTS: We mapped the dispersion of leprosy incidence in Brazil from 2014 to 2018, and found a particularly high number of cases in central Brazil in 2014 that further increased in 2018 in the state of Mato Grosso. For some municipalities, up to 80% of cases showed some data discrepancy. Of a total of 21,047 discrepancies detected, the most common was "operational classification does not match the clinical form." After data processing, we identified a total of 77,628 cases with missing data. The sensitivity and specificity of the AI model applied for the operational classification of leprosy was 93.97% and 87.09%, respectively. CONCLUSIONS: The proposed app was able to recognize patterns in leprosy cases registered in the SINAN database and to classify new patients with paucibacillary or multibacillary leprosy, thereby reducing the probability of incorrect assignment by health centers. The collection and notification of data on leprosy in Brazil seem to lack specific validation to increase the quality of the data for implementations via AI. The AI models implemented in this work had satisfactory accuracy across Brazilian states and could be a complementary diagnosis tool, especially in remote areas with few specialist physicians.

Insights into Computational Drug Repurposing for Neurodegenerative Disease.

Computational drug repurposing has the ability to remarkably reduce drug development time and cost in an era where these factors are prohibitively high. Several examples of successful repurposed drugs exist in fields such as oncology, diabetes, leprosy, inflammatory bowel disease, among others, however computational drug repurposing in neurodegenerative disease has presented several unique challenges stemming from the lack of validation methods and difficulty in studying heterogenous diseases of aging. Here, we examine existing approaches to computational drug repurposing, including molecular, clinical, and biophysical methods, and propose data sources and methods to advance computational drug repurposing in neurodegenerative disease using Alzheimer's disease as an example.

A novel integrated molecular and serological analysis method to predict new cases of leprosy amongst household contacts.

BACKGROUND: Early detection of Mycobacterium leprae is a key strategy for disrupting the transmission chain of leprosy and preventing the potential onset of physical disabilities. Clinical diagnosis is essential, but some of the presented symptoms may go unnoticed, even by specialists. In areas of greater endemicity, serological and molecular tests have been performed and analyzed separately for the follow-up of household contacts, who are at high risk of developing the disease. The accuracy of these tests is still debated, and it is necessary to make them more reliable, especially for the identification of cases of leprosy between contacts. We proposed an integrated analysis of molecular and serological methods using artificial intelligence by the random forest (RF) algorithm to better diagnose and predict new cases of leprosy. METHODS: The study was developed in Governador Valadares, Brazil, a hyperendemic region for leprosy. A longitudinal study was performed, including new cases diagnosed in 2011 and their respective household contacts, who were followed in 2011, 2012, and 2016. All contacts were diligently evaluated by clinicians from Reference Center for Endemic Diseases (CREDEN-PES) before being classified as asymptomatic. Samples of slit skin smears (SSS) from the earlobe of the patients and household contacts were collected for quantitative polymerase chain reaction (qPCR) of 16S rRNA, and peripheral blood samples were collected for ELISA assays to detect LID-1 and ND-O-LID. RESULTS: The statistical analysis of the tests revealed sensitivity for anti-LID-1 (63.2%), anti-ND-O-LID (57.9%), qPCR SSS (36.8%), and smear microscopy (30.2%). However, the use of RF allowed for an expressive increase in sensitivity in the diagnosis of multibacillary leprosy (90.5%) and especially paucibacillary leprosy (70.6%). It is important to report that the specificity was 92.5%. CONCLUSION: The proposed model using RF allows for the diagnosis of leprosy with high sensitivity and specificity and the early identification of new cases among household contacts.

Bi-PSSM: Position specific scoring matrix based intelligent computational model for identification of mycobacterial membrane proteins.

Mycobacterium is a pathogenic bacterium, which is a causative agent of tuberculosis (TB) and leprosy. These diseases are very crucial and become the cause of death of millions of people every year in the world. So, the characterize structure of membrane proteins of the protozoan play a vital role in the field of drug discovery because, without any knowledge about this Mycobacterium's membrane protein and their types, the scientists are unable to treat this pathogenic protozoan. So, an accurate and competitive computational model is needed to characterize this uncharacterized structure of mycobacterium. Series of attempts were carried out in this connection. Split amino acid compositions, Unbiased-Dipeptide peptide compositions (Unb-DPC), Over-represented tri-peptide compositions, compositions & translation were the few recent encoding techniques followed by different researchers in their publications. Although considerable results have been achieved by these models, still there is a gap which is filled in this study. In this study, an evolutionary feature extraction technique position specific scoring matrix (PSSM) is applied in order to extract evolutionary information from protein sequences. Consequently, 99.6% accuracy was achieved by the learning algorithms. The experimental results demonstrated that the proposed computational model will lead to develop a powerful tool for anti-mycobacterium drugs as well as play a promising rule in proteomic and bioinformatics.

Support vector machines for temporal classification of block design fMRI data.

This paper treats support vector machine (SVM) classification applied to block design fMRI, extending our previous work with linear discriminant analysis [LaConte, S., Anderson, J., Muley, S., Ashe, J., Frutiger, S., Rehm, K., Hansen, L.K., Yacoub, E., Hu, X., Rottenberg, D., Strother S., 2003a. The evaluation of preprocessing choices in single-subject BOLD fMRI using NPAIRS performance metrics. NeuroImage 18, 10-27; Strother, S.C., Anderson, J., Hansen, L.K., Kjems, U., Kustra, R., Siditis, J., Frutiger, S., Muley, S., LaConte, S., Rottenberg, D. 2002. The quantitative evaluation of functional neuroimaging experiments: the NPAIRS data analysis framework. NeuroImage 15, 747-771]. We compare SVM to canonical variates analysis (CVA) by examining the relative sensitivity of each method to ten combinations of preprocessing choices consisting of spatial smoothing, temporal detrending, and motion correction. Important to the discussion are the issues of classification performance, model interpretation, and validation in the context of fMRI. As the SVM has many unique properties, we examine the interpretation of support vector models with respect to neuroimaging data. We propose four methods for extracting activation maps from SVM models, and we examine one of these in detail. For both CVA and SVM, we have classified individual time samples of whole brain data, with TRs of roughly 4 s, thirty slices, and nearly 30,000 brain voxels, with no averaging of scans or prior feature selection.

Measurement of retinal vessel widths from fundus images based on 2-D modeling.

Changes in retinal vessel diameter are an important sign of diseases such as hypertension, arteriosclerosis and diabetes mellitus. Obtaining precise measurements of vascular widths is a critical and demanding process in automated retinal image analysis as the typical vessel is only a few pixels wide. This paper presents an algorithm to measure the vessel diameter to subpixel accuracy. The diameter measurement is based on a two-dimensional difference of Gaussian model, which is optimized to fit a two-dimensional intensity vessel segment. The performance of the method is evaluated against Brinchmann-Hansen's half height, Gregson's rectangular profile and Zhou's Gaussian model. Results from 100 sample profiles show that the presented algorithm is over 30% more precise than the compared techniques and is accurate to a third of a pixel.

Optimization of rifamycin B fermentation in shake flasks via a machine-learning-based approach.

Rifamycin B is an important polyketide antibiotic used in the treatment of tuberculosis and leprosy. We present results on medium optimization for Rifamycin B production via a barbital insensitive mutant strain of Amycolatopsis mediterranei S699. Machine-learning approaches such as Genetic algorithm (GA), Neighborhood analysis (NA) and Decision Tree technique (DT) were explored for optimizing the medium composition. Genetic algorithm was applied as a global search algorithm while NA was used for a guided local search and to develop medium predictors. The fermentation medium for Rifamycin B consisted of nine components. A large number of distinct medium compositions are possible by variation of concentration of each component. This presents a large combinatorial search space. Optimization was achieved within five generations via GA as well as NA. These five generations consisted of 178 shake-flask experiments, which is a small fraction of the search space. We detected multiple optima in the form of 11 distinct medium combinations. These medium combinations provided over 600% improvement in Rifamycin B productivity. Genetic algorithm performed better in optimizing fermentation medium as compared to NA. The Decision Tree technique revealed the media-media interactions qualitatively in the form of sets of rules for medium composition that give high as well as low productivity.