The role of machine learning in advancing diabetic foot: a review

Frontiers in Endocrinology, Journal Year: 2024, Volume and Issue: 15

Published: April 29, 2024

Background Diabetic foot complications impose a significant strain on healthcare systems worldwide, acting as principal cause of morbidity and mortality in individuals with diabetes mellitus. While traditional methods diagnosing treating these conditions have faced limitations, the emergence Machine Learning (ML) technologies heralds new era, offering promise revolutionizing diabetic care through enhanced precision tailored treatment strategies. Objective This review aims to explore transformative impact ML managing complications, highlighting its potential advance diagnostic accuracy therapeutic approaches by leveraging developments medical imaging, biomarker detection, clinical biomechanics. Methods A meticulous literature search was executed across PubMed, Scopus, Google Scholar databases identify pertinent articles published up March 2024. The strategy carefully crafted, employing combination keywords such “Machine Learning,” “Diabetic Foot,” Foot Ulcers,” Care,” “Artificial Intelligence,” “Predictive Modeling.” offers an in-depth analysis foundational principles algorithms that constitute ML, placing special emphasis their relevance sciences, particularly within specialized domain pathology. Through incorporation illustrative case studies schematic diagrams, endeavors elucidate intricate computational methodologies involved. Results has proven be invaluable deriving critical insights from complex datasets, enhancing both planning for management. highlights efficacy decision-making, underscored comparative analyses prognostic assessments applications care. Conclusion culminates prospective assessment trajectory realm We believe despite challenges limitations ethical considerations, remains at forefront paradigms management are globally applicable precision-oriented. technological evolution unprecedented possibilities opportunities patient

Language: Английский

---

Artificial Intelligence Methodologies Applied to Technologies for Screening, Diagnosis and Care of the Diabetic Foot: A Narrative Review

Biosensors, Journal Year: 2022, Volume and Issue: 12(11), P. 985 - 985

Published: Nov. 8, 2022

Diabetic foot syndrome is a multifactorial pathology with at least three main etiological factors, i.e., peripheral neuropathy, arterial disease, and infection. In addition to complexity, another distinctive trait of diabetic its insidiousness, due frequent lack early symptoms. recent years, it has become clear that the prevalence increasing, among diabetes complications stronger impact on patient's quality life. Considering complex nature this syndrome, artificial intelligence (AI) methodologies appear adequate address aspects such as timely screening for identification risk ulcers (or, even worse, amputation), based appropriate sensor technologies. review, we summarize findings pertinent studies in field, paying attention both AI-based methodological physiological/clinical study outcomes. The analyzed show AI application data derived by different technologies provides promising results, but our opinion future may benefit from inclusion quantitative measures simple sensors, which are still scarcely exploited.

Language: Английский

---

The impact of machine learning on the prediction of diabetic foot ulcers – A systematic review

Journal of Tissue Viability, Journal Year: 2024, Volume and Issue: unknown

Published: July 1, 2024

Globally, diabetes mellitus poses a significant health challenge as well the associated complications of diabetes, such diabetic foot ulcers (DFUs). The early detection DFUs is important in healing process and machine learning may be able to help inform clinical staff during treatment process.

Language: Английский

---

Gait acceleration-based diabetes detection using hybrid deep learning

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 92, P. 105998 - 105998

Published: Feb. 8, 2024

Diabetes is a medical condition affecting multiple organs and systems due to high blood glucose levels in the body which cause diabetic neuropathy foot ulcers. Conventionally, diabetes detected using invasive methods such as pricking finger measuring glucose. However, are not convenient can pain patients. An alternative method detect use gait analysis abnormalities be analysed patterns predict severity. To our best knowledge, no studies have investigated of acceleration for detection hybrid deep learning models. Current research utilises models with non-gait data electrocardiography Pima Indians Database. This paper aims classify by utilising from wearable sensors placed on hip, knees, ankles, employing model CNN-LSTM. The proposed CNN-LSTM consists two convolutional layers LSTM layers. By combining models, extract important features learn classification. performance compared CNN accuracy, precision, recall, F1 score, AUC ROC. Compared existing methods, has achieved higher accuracy 91.25%, surpassing that current methods. Hence, this demonstrates non-invasive techniques hold potential replace traditional In future, muscle activation forces together improve detection.

Language: Английский

---

Artificial Intelligence-Based Digital Biomarkers for Type 2 Diabetes: A Review

Canadian Journal of Cardiology, Journal Year: 2024, Volume and Issue: 40(10), P. 1922 - 1933

Published: Aug. 5, 2024

Type 2 diabetes mellitus (T2DM), a complex metabolic disorder that burdens the health care system, requires early detection and treatment. Recent strides in digital technologies, coupled with artificial intelligence (AI), may have potential to revolutionize T2DM screening, diagnosis of complications, management through development biomarkers. This review provides an overview applications AI-driven biomarkers context diagnosing managing patients T2DM. The benefits using multisensor devices develop are discussed. summary these findings patterns between model architecture sensor type presented. In addition, we highlight pivotal role AI techniques clinical intervention implementation, encompassing decision support systems, telemedicine interventions, population initiatives. Challenges such as data privacy, algorithm interpretability, regulatory considerations also highlighted, alongside future research directions explore use screening management.

Language: Английский

---

sEMG Signal-Based Lower Limb Movements Recognition Using Tunable Q-Factor Wavelet Transform and Kraskov Entropy

IRBM, Journal Year: 2023, Volume and Issue: 44(4), P. 100773 - 100773

Published: Feb. 28, 2023

Language: Английский

---

A Machine Learning-Based Severity Prediction Tool for the Michigan Neuropathy Screening Instrument

Diagnostics, Journal Year: 2023, Volume and Issue: 13(2), P. 264 - 264

Published: Jan. 11, 2023

Diabetic sensorimotor polyneuropathy (DSPN) is a serious long-term complication of diabetes, which may lead to foot ulceration and amputation. Among the screening tools for DSPN, Michigan neuropathy instrument (MNSI) frequently deployed, but it lacks straightforward rating severity. A DSPN severity grading system has been built simulated MNSI, utilizing longitudinal data captured over 19 years from Epidemiology Diabetes Interventions Complications (EDIC) trial. Machine learning algorithms were used establish MNSI factors patient outcomes characterise features with best ability detect nomogram based on multivariable logistic regression was designed, developed validated. The extra tree model applied identify top seven ranked that identified namely vibration perception (R), 10-gm filament, previous diabetic neuropathy, (L), presence callus, deformities fissure. nomogram’s area under curve (AUC) 0.9421 0.946 internal external datasets, respectively. probability predicted created using score. An independent dataset validate model’s performance. patients divided into four different levels, i.e., absent, mild, moderate, severe, cut-off values 10.50, 12.70 15.00 less than 50, 75 100%, We provide an easy-to-use, reproducible approach determine prognosis in DSPN.

Language: Английский

---

Wearable Movement Exploration Device with Machine Learning Algorithm for Screening and Tracking Diabetic Neuropathy—A Cross-Sectional, Diagnostic, Comparative Study

Biosensors, Journal Year: 2024, Volume and Issue: 14(4), P. 166 - 166

Published: March 29, 2024

Background: Diabetic neuropathy is one of the most common complications diabetes mellitus. The aim this study to evaluate Moveo device, a novel device that uses machine learning (ML) algorithm detect and track diabetic neuropathy. comprises 4 sensors positioned on back hands feet accompanied by mobile application gathers data ML algorithms are hosted cloud platform. measure movement signals, which then transferred through application. triggers pipeline for feature extraction subsequently feeds model with these extracted features. Methods: pilot included 23 participants. Eleven patients suspected were in experimental group. In control group, 8 had radiculopathy, participants healthy. All underwent an electrodiagnostic examination (EDx) examination, consists placed participant’s use participant performs six tests part standard neurological calculates probability A user experience questionnaire was used compare experiences regard both methods. Results: total accuracy 82.1%, 78% sensitivity 87% specificity. high linear correlation up 0.722 observed between EDx features, underpins model’s adequacy. revealed majority preferred less painful method. Conclusions: represents accurate, easy-to-use suitable home environments, showing promising results potential future usage.

Language: Английский

---

SignAPROS: An integrated hardware and software system for acquisition, processing, and analysis of bio-signals

Software Impacts, Journal Year: 2025, Volume and Issue: unknown, P. 100741 - 100741

Published: March 1, 2025

Language: Английский

---

Deep Learning-Based Classification and Feature Extraction for Predicting Pathogenesis of Foot Ulcers in Patients with Diabetes

Diagnostics, Journal Year: 2023, Volume and Issue: 13(12), P. 1983 - 1983

Published: June 6, 2023

The World Health Organization (WHO) has identified that diabetes mellitus (DM) is one of the most prevalent disease worldwide. Individuals with DM have a higher risk mortality, and it crucial to prioritize treatment foot ulcers, which significant complication associated disease, as they lead development plantar results in need amputate part or leg. People are at experiencing various complications, such heart eye problems, kidney dysfunction, nerve damage, skin issues, dental diseases. Unawareness diabetic ulcers (DFU) contributing factor mortality patients. Evolving technological advancements deep learning techniques can be used predict symptoms early possible, helps provide effective This research introduces methodology for analyzing images patients, focusing on feature extraction classification. dataset this study was collected from historical medical records patients diabetes, who commonly experience major complication. pre-processed segmented, features were extracted using recurrent neural network (DRNN). Image numerical/text data separately, normal abnormal ranges identified. Foot separated classified pre-trained fast convolutional (PFCNN) U++net. classification procedure involves analysis their pathogenesis. To assess effectiveness proposed technique, presented simulation results, including confusion matrix receiver operating characteristic curve. These specifically focused predicting two classes: ulcerations. yielded parameters, accuracy, precision, recall curve, area under main goal introduce an novel technique assessing ulceration leveraging ulcer images. researchers segmented data. They then extract based numerical text U++net examine forecast (DFU). assessed accuracy 99.32% by simulating ulcers. A comparison made between existing approaches.

Language: Английский

---