The field of optimization is focused on the formulation, analysis, and resolution of problems involving the minimization or maximization of functions. A particular subclass of optimization problems, known as empirical risk minimization, involves fitting a model to observed data. These problems play a central role in various areas such as machine learning, statistical modeling, and decision theory, where the objective is to find a model that best approximates underlying patterns in the data by minimizing a specified loss or risk function. In deep learning (DL) systems, various optimization algorithms are utilized with the gradient descent (GD) algorithm being one of the most significant and effective. Research studies have improved the GD algorithm and developed various successful variants, including stochastic gradient descent (SGD) with momentum, AdaGrad, RMSProp, and Adam. This article provides a comparative analysis of these stochastic gradient descent algorithms based on their accuracy, loss, and training time, as well as the loss of each algorithm in generating an optimization solution. Experiments were conducted using Transfer Learning (TL) technique based on the pre-trained ResNet50 base model for image classification, with a focus on stochastic gradient (SG) for performances optimization. The case study in this work is based on a data extract from the Alzheimer's image dataset, which contains four classes such as Mild Demented, Moderate Demented, Non-Demented, and Very Mild Demented. The obtained results with the Adam and SGD momentum optimizers achieved the highest accuracy of 97.66 % and 97.58 %, respectively.
{"title":"Optimizing ResNet50 performance using stochastic gradient descent on MRI images for Alzheimer's disease classification","authors":"Mohamed Amine Mahjoubi , Driss Lamrani , Shawki Saleh , Wassima Moutaouakil , Asmae Ouhmida , Soufiane Hamida , Bouchaib Cherradi , Abdelhadi Raihani","doi":"10.1016/j.ibmed.2025.100219","DOIUrl":"10.1016/j.ibmed.2025.100219","url":null,"abstract":"<div><div>The field of optimization is focused on the formulation, analysis, and resolution of problems involving the minimization or maximization of functions. A particular subclass of optimization problems, known as empirical risk minimization, involves fitting a model to observed data. These problems play a central role in various areas such as machine learning, statistical modeling, and decision theory, where the objective is to find a model that best approximates underlying patterns in the data by minimizing a specified loss or risk function. In deep learning (DL) systems, various optimization algorithms are utilized with the gradient descent (GD) algorithm being one of the most significant and effective. Research studies have improved the GD algorithm and developed various successful variants, including stochastic gradient descent (SGD) with momentum, AdaGrad, RMSProp, and Adam. This article provides a comparative analysis of these stochastic gradient descent algorithms based on their accuracy, loss, and training time, as well as the loss of each algorithm in generating an optimization solution. Experiments were conducted using Transfer Learning (TL) technique based on the pre-trained ResNet50 base model for image classification, with a focus on stochastic gradient (SG) for performances optimization. The case study in this work is based on a data extract from the Alzheimer's image dataset, which contains four classes such as Mild Demented, Moderate Demented, Non-Demented, and Very Mild Demented. The obtained results with the Adam and SGD momentum optimizers achieved the highest accuracy of 97.66 % and 97.58 %, respectively.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"11 ","pages":"Article 100219"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143173637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study was conducted due to the lack of applications that can assist people intreating common external wounds. Therefore, we proposed the application of image-based detection which takes external wounds and identifies them using Artificial Intelligence namely LukaKu. In addition to detecting the type of wound that occurs, the application is expected to be able to produce first aid and medicine for each existing external wound label. The model used is YOLOv5 with various versions, namely YOLOv5n, YOLOv5s, YOLOv5m, YOLOv5l, and YOLOv5x. By calculating the validation data, each version has its own precision, recall, f1-score, and Mean Average Precision (mAP) values which are the comparison factors in determining the best model version, where YOLOv5l with mAP value of 0.785 is the best result and YOLOv5n with mAP value of 0.588 is the result with the lowest value. In the model development process, datasets of external injuries are needed to be used during the training process and test datasets for each existing model version. After each version of the model has been successfully built and analysed, the model with the best value is implemented in the mobile application, making it easier for users to access.
这项研究是由于缺乏应用程序，可以帮助人们治疗常见的外部伤口。因此，我们提出了基于图像的检测应用，该检测采用人工智能即LukaKu来识别外部伤口。除了检测发生的伤口类型之外，该应用程序预计能够为每个现有的外部伤口标签生产急救和药物。型号为YOLOv5，有YOLOv5n、YOLOv5s、YOLOv5m、YOLOv5l、YOLOv5x等多个版本。通过计算验证数据，每个版本都有自己的精度、召回率、f1-score和Mean Average precision （mAP）值，这些值是确定最佳模型版本的比较因素，其中mAP值为0.785的YOLOv5l为最佳结果，mAP值为0.588的YOLOv5n为最低结果。在模型开发过程中，在训练过程中需要使用外伤性数据集，在现有的各个模型版本中需要使用测试数据集。在成功构建和分析了每个版本的模型后，将最有价值的模型实现在移动应用程序中，使用户更容易访问。
{"title":"A mobile application LukaKu as a tool for detecting external wounds with artificial intelligence","authors":"Dessy Novita , Herika Hayurani , Eva Krishna Sutedja , Firdaus Ryan Pratomo , Achmad Dino Saputra , Zahra Ramadhanti , Nuryadin Abutani , Muhammad Rafi Triandi , Aldin Mubarok Guferol , Anindya Apriliyanti Pravitasari , Fajar Wira Adikusuma , Atiek Rostika Noviyanti","doi":"10.1016/j.ibmed.2025.100200","DOIUrl":"10.1016/j.ibmed.2025.100200","url":null,"abstract":"<div><div>This study was conducted due to the lack of applications that can assist people intreating common external wounds. Therefore, we proposed the application of image-based detection which takes external wounds and identifies them using Artificial Intelligence namely LukaKu. In addition to detecting the type of wound that occurs, the application is expected to be able to produce first aid and medicine for each existing external wound label. The model used is YOLOv5 with various versions, namely YOLOv5n, YOLOv5s, YOLOv5m, YOLOv5l, and YOLOv5x. By calculating the validation data, each version has its own precision, recall, f1-score, and Mean Average Precision (mAP) values which are the comparison factors in determining the best model version, where YOLOv5l with mAP value of 0.785 is the best result and YOLOv5n with mAP value of 0.588 is the result with the lowest value. In the model development process, datasets of external injuries are needed to be used during the training process and test datasets for each existing model version. After each version of the model has been successfully built and analysed, the model with the best value is implemented in the mobile application, making it easier for users to access.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"11 ","pages":"Article 100200"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143174331","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.ibmed.2025.100232
Maria Jiménez , Cristian Soza-Ried , Vasko Kramer , Sebastian A. Ríos , Arlette Haeger , Carlos Juri , Horacio Amaral , Pedro Chana-Cuevas
Parkinsonian syndrome (PS) is characterized by bradykinesia, resting tremor, rigidity, and encapsulates the clinical manifestation observed in various neurodegenerative disorders. Positron emission tomography (PET) imaging plays an important role in diagnosing PS by detecting the progressive loss of dopaminergic neurons. This study aimed to develop and compare five machine-learning models for the automatic classification of 204 [18F]PR04.MZ PET images, distinguishing between patients with PS and subjects without clinical evidence for dopaminergic deficit (SWEDD). Previously analyzed and classified by three expert blind readers into PS compatible (1) and SWEDDs (0), the dataset was processed in both two-dimensional and three-dimensional formats. Five widely used pattern recognition algorithms were trained and validated their performance. These algorithms were compared against the majority reading of expert diagnosis, considered the gold standard. Comparing the accuracy of 2D and 3D format images suggests that, without the depth dimension, a single image may overemphasize specific regions. Overall, three models outperformed with an accuracy greater than 98 %, demonstrating that machine-learning models trained with [18F]PR04.MZ PET images can provide a highly accurate and precise tool to support clinicians in automatic PET image analysis. This approach may be a first step in reducing the time required for interpretation, as well as increase certainty in the diagnostic process.
帕金森综合征（Parkinsonian Syndrome，PS）以运动迟缓、静止性震颤和僵直为特征，是各种神经退行性疾病的临床表现。正电子发射断层扫描（PET）成像通过检测多巴胺能神经元的逐渐丧失，在诊断帕金森综合征中发挥着重要作用。本研究旨在开发和比较五种机器学习模型，用于对204张[18F]PR04.MZ PET图像进行自动分类，区分PS患者和无多巴胺能缺失临床证据的受试者（SWEDD）。该数据集之前由三位盲人专家进行了分析和分类，分为 PS 相容性（1）和 SWEDD（0），并以二维和三维格式进行了处理。对五种广泛使用的模式识别算法进行了训练，并对其性能进行了验证。这些算法与被视为金标准的专家诊断的多数读数进行了比较。比较二维和三维格式图像的准确性表明，如果没有深度维度，单一图像可能会过分强调特定区域。总体而言，三个模型的准确率都超过了 98%，这表明使用[18F]PR04.MZ PET 图像训练的机器学习模型可以提供一种高度准确和精确的工具，为临床医生自动 PET 图像分析提供支持。这种方法可能是减少判读所需时间的第一步，并能提高诊断过程的确定性。
{"title":"Image-based machine learning model as a tool for classification of [18F]PR04.MZ PET images in patients with parkinsonian syndrome","authors":"Maria Jiménez , Cristian Soza-Ried , Vasko Kramer , Sebastian A. Ríos , Arlette Haeger , Carlos Juri , Horacio Amaral , Pedro Chana-Cuevas","doi":"10.1016/j.ibmed.2025.100232","DOIUrl":"10.1016/j.ibmed.2025.100232","url":null,"abstract":"<div><div>Parkinsonian syndrome (PS) is characterized by bradykinesia, resting tremor, rigidity, and encapsulates the clinical manifestation observed in various neurodegenerative disorders. Positron emission tomography (PET) imaging plays an important role in diagnosing PS by detecting the progressive loss of dopaminergic neurons. This study aimed to develop and compare five machine-learning models for the automatic classification of 204 [<sup>18</sup>F]PR04.MZ PET images, distinguishing between patients with PS and subjects without clinical evidence for dopaminergic deficit (SWEDD). Previously analyzed and classified by three expert blind readers into PS compatible (1) and SWEDDs (0), the dataset was processed in both two-dimensional and three-dimensional formats. Five widely used pattern recognition algorithms were trained and validated their performance. These algorithms were compared against the majority reading of expert diagnosis, considered the gold standard. Comparing the accuracy of 2D and 3D format images suggests that, without the depth dimension, a single image may overemphasize specific regions. Overall, three models outperformed with an accuracy greater than 98 %, demonstrating that machine-learning models trained with [<sup>18</sup>F]PR04.MZ PET images can provide a highly accurate and precise tool to support clinicians in automatic PET image analysis. This approach may be a first step in reducing the time required for interpretation, as well as increase certainty in the diagnostic process.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"11 ","pages":"Article 100232"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143628516","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.ibmed.2025.100235
Alexander J. Idarraga , David F. Schneider
Objective
Telehealth is an increasingly important method for delivering care. Health systems lack the ability to accurately predict which telehealth visits will fail due to poor connection, poor technical literacy, or other reasons. This results in wasted resources and disrupted patient care. The purpose of this study is to characterize and compare various methods for predicting telehealth visit failure, and to determine the prediction method most suited for implementation in a real-time operational setting.
Methods
A single-center, retrospective cohort study was conducted using data sourced from our data warehouse. Patient demographic information and data characterizing prior visit success and engagement with electronic health tools were included. Three main model types were evaluated: an existing scoring model developed by Hughes et al., a regression-based scoring model, and Machine Learning classifiers. Variables were selected for their importance and anticipated availability; Number Needed to Treat was used to demonstrate the number of interventions (e.g. pre-visit phone calls) required to improve success rates in the context of weekly patient volumes.
Results
217, 229 visits spanning 480 days were evaluated, of which 22,443 (10.33 %) met criteria for failure. Hughes et al.’s model applied to our data yielded an Area Under the Receiver Operating Characteristics Curve (AUC ROC) of 0.678 when predicting failure. A score-based model achieved an AUC ROC of 0.698. Logistic Regression, Random Forest, and Gradient Boosting models demonstrated AUC ROCs ranging from 0.7877 to 0.7969. A NNT of 32 was achieved if the 263 highest-risk patients were selected in a low-volume week using the RF classifier, compared to an expected NNT of 90 if the same number of patients were randomly selected.
Conclusions
Machine Learning classifiers demonstrated superiority over score-based methods for predicting telehealth visit failure. Prospective evaluation is required; evaluation using NNT as a metric can help to operationalize these models.
{"title":"A comparison of techniques for predicting telehealth visit failure","authors":"Alexander J. Idarraga , David F. Schneider","doi":"10.1016/j.ibmed.2025.100235","DOIUrl":"10.1016/j.ibmed.2025.100235","url":null,"abstract":"<div><h3>Objective</h3><div>Telehealth is an increasingly important method for delivering care. Health systems lack the ability to accurately predict which telehealth visits will fail due to poor connection, poor technical literacy, or other reasons. This results in wasted resources and disrupted patient care. The purpose of this study is to characterize and compare various methods for predicting telehealth visit failure, and to determine the prediction method most suited for implementation in a real-time operational setting.</div></div><div><h3>Methods</h3><div>A single-center, retrospective cohort study was conducted using data sourced from our data warehouse. Patient demographic information and data characterizing prior visit success and engagement with electronic health tools were included. Three main model types were evaluated: an existing scoring model developed by Hughes et al., a regression-based scoring model, and Machine Learning classifiers. Variables were selected for their importance and anticipated availability; Number Needed to Treat was used to demonstrate the number of interventions (e.g. pre-visit phone calls) required to improve success rates in the context of weekly patient volumes.</div></div><div><h3>Results</h3><div>217, 229 visits spanning 480 days were evaluated, of which 22,443 (10.33 %) met criteria for failure. Hughes et al.’s model applied to our data yielded an Area Under the Receiver Operating Characteristics Curve (AUC ROC) of 0.678 when predicting failure. A score-based model achieved an AUC ROC of 0.698. Logistic Regression, Random Forest, and Gradient Boosting models demonstrated AUC ROCs ranging from 0.7877 to 0.7969. A NNT of 32 was achieved if the 263 highest-risk patients were selected in a low-volume week using the RF classifier, compared to an expected NNT of 90 if the same number of patients were randomly selected.</div></div><div><h3>Conclusions</h3><div>Machine Learning classifiers demonstrated superiority over score-based methods for predicting telehealth visit failure. Prospective evaluation is required; evaluation using NNT as a metric can help to operationalize these models.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"11 ","pages":"Article 100235"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143747628","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.ibmed.2025.100220
Esaie Naroum , Ebenezer Maka Maka , Hamadjam Abboubakar , Paul Dayang , Appolinaire Batoure Bamana , Benjamin Garga , Hassana Daouda Daouda , Mohsen Bakouri , Ilyas Khan
The Plasmodium parasite, which causes malaria is transmitted by Anopheles mosquitoes, and remains a major development barrier in Africa. This is particularly true considering the conducive environment that promotes the spread of malaria. This study examines several machine learning approaches, such as long short term memory (LSTM), random forests (RF), support vector machines (SVM), and data regularization models including Ridge, Lasso, and ElasticNet, in order to forecast the occurrence of malaria in the Adamaoua region of Cameroon. The LSTM, a recurrent neural network variant, performed the best with 76% accuracy and a low error rate (RMSE = 0.08). Statistical evidence indicates that temperatures exceeding 34 degrees halt mosquito vector reproduction, thereby slowing the spread of malaria. However, humidity increases the morbidity of the condition. The survey also identified high-risk areas in Ngaoundéré Rural and Urban and Meiganga. Between 2018 and 2022, the Adamaoua region had 20.1%, 12.3%, and 10.0% of malaria cases, respectively, in these locations. According to the estimate, the number of malaria cases in the Adamaoua region will rise gradually between 2023 and 2026, peaking in 2029 before declining in 2031.
{"title":"Comparative analysis of deep learning and machine learning techniques for forecasting new malaria cases in Cameroon’s Adamaoua region","authors":"Esaie Naroum , Ebenezer Maka Maka , Hamadjam Abboubakar , Paul Dayang , Appolinaire Batoure Bamana , Benjamin Garga , Hassana Daouda Daouda , Mohsen Bakouri , Ilyas Khan","doi":"10.1016/j.ibmed.2025.100220","DOIUrl":"10.1016/j.ibmed.2025.100220","url":null,"abstract":"<div><div>The Plasmodium parasite, which causes malaria is transmitted by Anopheles mosquitoes, and remains a major development barrier in Africa. This is particularly true considering the conducive environment that promotes the spread of malaria. This study examines several machine learning approaches, such as long short term memory (LSTM), random forests (RF), support vector machines (SVM), and data regularization models including Ridge, Lasso, and ElasticNet, in order to forecast the occurrence of malaria in the Adamaoua region of Cameroon. The LSTM, a recurrent neural network variant, performed the best with 76% accuracy and a low error rate (RMSE = 0.08). Statistical evidence indicates that temperatures exceeding 34 degrees halt mosquito vector reproduction, thereby slowing the spread of malaria. However, humidity increases the morbidity of the condition. The survey also identified high-risk areas in Ngaoundéré Rural and Urban and Meiganga. Between 2018 and 2022, the Adamaoua region had 20.1%, 12.3%, and 10.0% of malaria cases, respectively, in these locations. According to the estimate, the number of malaria cases in the Adamaoua region will rise gradually between 2023 and 2026, peaking in 2029 before declining in 2031.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"11 ","pages":"Article 100220"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143388305","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.ibmed.2025.100248
ArunaDevi Karuppasamy , Hamza zidoum , Majda Said Sultan Al-Rashdi , Maiya Al-Bahri
The Deep Learning (DL) has demonstrated a significant impact on a various pattern recognition applications, resulting in significant advancements in areas such as visual recognition, autonomous cars, language processing, and healthcare. Nowadays, deep learning was widely applied on the medical images to identify the diseases efficiently. Still, the use of applications in clinical settings is now limited to a small number. The main factors to this might be due to an inadequate annotated data, noises in the images and challenges related to collecting data. Our research proposed a convolutional autoencoder to classify the breast cancer tumors, using the Sultan Qaboos University Hospital(SQUH) and BreakHis datasets. The proposed model named Convolutional AutoEncoder with modified Loss Function (CAE-LF) achieved a good performance, by attaining a F1-score of 0.90, recall of 0.89, and accuracy of 91%. The results obtained are comparable to those obtained in earlier researches. Additional analyses conducted on the SQUH dataset demonstrate that it yields a good performance with an F1-score of 0.91, 0.93, 0.92, and 0.93 for 4x, 10x, 20x, and 40x magnifications, respectively. Our study highlights the potential of deep learning in analyzing medical images to classify breast tumors.
{"title":"Optimizing breast cancer diagnosis with convolutional autoencoders: Enhanced performance through modified loss functions","authors":"ArunaDevi Karuppasamy , Hamza zidoum , Majda Said Sultan Al-Rashdi , Maiya Al-Bahri","doi":"10.1016/j.ibmed.2025.100248","DOIUrl":"10.1016/j.ibmed.2025.100248","url":null,"abstract":"<div><div>The Deep Learning (DL) has demonstrated a significant impact on a various pattern recognition applications, resulting in significant advancements in areas such as visual recognition, autonomous cars, language processing, and healthcare. Nowadays, deep learning was widely applied on the medical images to identify the diseases efficiently. Still, the use of applications in clinical settings is now limited to a small number. The main factors to this might be due to an inadequate annotated data, noises in the images and challenges related to collecting data. Our research proposed a convolutional autoencoder to classify the breast cancer tumors, using the Sultan Qaboos University Hospital(SQUH) and BreakHis datasets. The proposed model named Convolutional AutoEncoder with modified Loss Function (CAE-LF) achieved a good performance, by attaining a F1-score of 0.90, recall of 0.89, and accuracy of 91%. The results obtained are comparable to those obtained in earlier researches. Additional analyses conducted on the SQUH dataset demonstrate that it yields a good performance with an F1-score of 0.91, 0.93, 0.92, and 0.93 for 4x, 10x, 20x, and 40x magnifications, respectively. Our study highlights the potential of deep learning in analyzing medical images to classify breast tumors.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"11 ","pages":"Article 100248"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143887937","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.ibmed.2025.100267
Shuaibu Saidu Musa , Adamu Muhammad Ibrahim , Muhammad Yasir Alhassan , Abubakar Hafs Musa , Abdulrahman Garba Jibo , Auwal Rabiu Auwal , Olalekan John Okesanya , Zhinya Kawa Othman , Muhammad Sadiq Abubakar , Mohamed Mustaf Ahmed , Carina Joane V. Barroso , Abraham Fessehaye Sium , Manuel B. Garcia , James Brian Flores , Adamu Safiyanu Maikifi , M.B.N. Kouwenhoven , Don Eliseo Lucero-Prisno
The fusion of molecular-scale engineering in nanotechnology with machine learning (ML) analytics is reshaping the field of precision medicine. Nanoparticles enable ultrasensitive diagnostics, targeted drug and gene delivery, and high-resolution imaging, whereas ML models mine vast multimodal datasets to optimize nanoparticle design, enhance predictive accuracy, and personalize treatment in real-time. Recent breakthroughs include ML-guided formulations of lipid, polymeric, and inorganic carriers that cross biological barriers; AI-enhanced nanosensors that flag early disease from breath, sweat, or blood; and nanotheranostic agents that simultaneously track and treat tumors. Comparative insights into Retrieval-Augmented Generation and supervised learning pipelines reveal distinct advantages for nanodevice engineering across diverse data environments. An expanded focus on explainable AI tools, such as SHAP, LIME, Grad-CAM, and Integrated Gradients, highlights their role in enhancing transparency, trust, and interpretability in nano-enabled clinical decisions. A structured narrative review method was applied, and key ML model performances were synthesized to strengthen analytical clarity. Emerging biodegradable nanomaterials, autonomous micro-nanorobots, and hybrid lab-on-chip systems promise faster point-of-care decisions but raise pressing questions about data integrity, interpretability, scalability, regulation, ethics, and equitable access. Addressing these hurdles will require robust data standards, privacy safeguards, interdisciplinary R&D networks, and flexible approval pathways to translate bench advances into bedside benefits for patients. This review synthesizes the current landscape, critical challenges, and future directions at the intersection of nanotechnology and ML in precision medicine.
{"title":"Nanotechnology and machine learning: a promising confluence for the advancement of precision medicine","authors":"Shuaibu Saidu Musa , Adamu Muhammad Ibrahim , Muhammad Yasir Alhassan , Abubakar Hafs Musa , Abdulrahman Garba Jibo , Auwal Rabiu Auwal , Olalekan John Okesanya , Zhinya Kawa Othman , Muhammad Sadiq Abubakar , Mohamed Mustaf Ahmed , Carina Joane V. Barroso , Abraham Fessehaye Sium , Manuel B. Garcia , James Brian Flores , Adamu Safiyanu Maikifi , M.B.N. Kouwenhoven , Don Eliseo Lucero-Prisno","doi":"10.1016/j.ibmed.2025.100267","DOIUrl":"10.1016/j.ibmed.2025.100267","url":null,"abstract":"<div><div>The fusion of molecular-scale engineering in nanotechnology with machine learning (ML) analytics is reshaping the field of precision medicine. Nanoparticles enable ultrasensitive diagnostics, targeted drug and gene delivery, and high-resolution imaging, whereas ML models mine vast multimodal datasets to optimize nanoparticle design, enhance predictive accuracy, and personalize treatment in real-time. Recent breakthroughs include ML-guided formulations of lipid, polymeric, and inorganic carriers that cross biological barriers; AI-enhanced nanosensors that flag early disease from breath, sweat, or blood; and nanotheranostic agents that simultaneously track and treat tumors. Comparative insights into Retrieval-Augmented Generation and supervised learning pipelines reveal distinct advantages for nanodevice engineering across diverse data environments. An expanded focus on explainable AI tools, such as SHAP, LIME, Grad-CAM, and Integrated Gradients, highlights their role in enhancing transparency, trust, and interpretability in nano-enabled clinical decisions. A structured narrative review method was applied, and key ML model performances were synthesized to strengthen analytical clarity. Emerging biodegradable nanomaterials, autonomous micro-nanorobots, and hybrid lab-on-chip systems promise faster point-of-care decisions but raise pressing questions about data integrity, interpretability, scalability, regulation, ethics, and equitable access. Addressing these hurdles will require robust data standards, privacy safeguards, interdisciplinary R&D networks, and flexible approval pathways to translate bench advances into bedside benefits for patients. This review synthesizes the current landscape, critical challenges, and future directions at the intersection of nanotechnology and ML in precision medicine.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"12 ","pages":"Article 100267"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144271155","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.ibmed.2025.100265
Abedin Keshavarz, Amir Lakizadeh
Polypharmacy, or the concurrent use of multiple medications, increases the risk of adverse effects due to drug interactions. As polypharmacy becomes more prevalent, forecasting these interactions is essential in the pharmaceutical field. Due to the limitations of clinical trials in detecting rare side effects associated with polypharmacy, computational methods are being developed to model these adverse effects. This study introduces a method named PU-MLP, based on a Multi-Layer Perceptron, to predict side effects from drug combinations. This research utilizes advanced machine learning techniques to explore the connections between medications and their adverse effects. The approach consists of three key stages: first, it creates an optimal representation of each drug using a combination of a random forest classifier, Graph Neural Networks (GNNs), and dimensionality reduction techniques. Second, it employs Positive Unlabeled learning to address data uncertainty. Finally, a Multi-Layer Perceptron model is utilized to predict polypharmacy side effects. Performance evaluation using 5-fold cross-validation shows that the proposed method surpasses other approaches, achieving impressive scores of 0.99, 0.99, and 0.98 in AUPR, AUC, and F1 measures, respectively.
{"title":"PU-MLP: A PU-learning based method for polypharmacy side-effects detection based on multi-layer perceptron and feature extraction techniques","authors":"Abedin Keshavarz, Amir Lakizadeh","doi":"10.1016/j.ibmed.2025.100265","DOIUrl":"10.1016/j.ibmed.2025.100265","url":null,"abstract":"<div><div>Polypharmacy, or the concurrent use of multiple medications, increases the risk of adverse effects due to drug interactions. As polypharmacy becomes more prevalent, forecasting these interactions is essential in the pharmaceutical field. Due to the limitations of clinical trials in detecting rare side effects associated with polypharmacy, computational methods are being developed to model these adverse effects. This study introduces a method named PU-MLP, based on a Multi-Layer Perceptron, to predict side effects from drug combinations. This research utilizes advanced machine learning techniques to explore the connections between medications and their adverse effects. The approach consists of three key stages: first, it creates an optimal representation of each drug using a combination of a random forest classifier, Graph Neural Networks (GNNs), and dimensionality reduction techniques. Second, it employs Positive Unlabeled learning to address data uncertainty. Finally, a Multi-Layer Perceptron model is utilized to predict polypharmacy side effects. Performance evaluation using 5-fold cross-validation shows that the proposed method surpasses other approaches, achieving impressive scores of 0.99, 0.99, and 0.98 in AUPR, AUC, and F1 measures, respectively.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"12 ","pages":"Article 100265"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144220989","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Dental caries is one of the major dental issues that is common among many individuals. It leads to tooth loss and affects the tooth root, creating a need to automatically detect dental caries to reduce treatment costs and prevent its consequences. The Lightweight Caries Segmentation Network (LCSNet) proposed in this study detects the location of dental caries by applying pixel-wise segmentation to dental photographs taken with various Android phones. LCSNet utilizes a Dual Multiscale Residual (DMR) block in both the encoder and decoder, adapts transfer learning through a pre-trained InceptionV3 model at the bottleneck layer, and incorporates a Squeeze and Excitation block in the skip connection, effectively extracting spatial information even from images where 95 % of the background and only 5 % represent the area of interest. A new dataset was developed by gathering oral photographs of dental caries from two hospitals, with advanced augmentation techniques applied. The LCSNet architecture demonstrated an accuracy of 97.36 %, precision of 73.1 %, recall of 70.2 %, an F1-Score of 71.14 %, and an Intersection-over-Union (IoU) of 56.8 %. Expert dentists confirmed that the LCSNet model proposed in this in vivo study accurately segments the position and texture of dental caries. Both qualitative and quantitative performance analyses, along with comparative analyses of efficiency and computational requirements, were conducted with other deep learning models. The proposed model outperforms existing deep learning models and shows significant potential for integration into a smartphone application-based oral disease detection system, potentially replacing some conventional clinically adapted methods.
{"title":"LCSNet: Lightweight Caries Segmentation Network for the segmentation of dental caries using smartphone photographs","authors":"Radha R.C. , B.S. Raghavendra , Rishabh Kumar Hota , K.R. Vijayalakshmi , Seema Patil , A.V. Narasimhadhan","doi":"10.1016/j.ibmed.2025.100304","DOIUrl":"10.1016/j.ibmed.2025.100304","url":null,"abstract":"<div><div>Dental caries is one of the major dental issues that is common among many individuals. It leads to tooth loss and affects the tooth root, creating a need to automatically detect dental caries to reduce treatment costs and prevent its consequences. The Lightweight Caries Segmentation Network (LCSNet) proposed in this study detects the location of dental caries by applying pixel-wise segmentation to dental photographs taken with various Android phones. LCSNet utilizes a Dual Multiscale Residual (DMR) block in both the encoder and decoder, adapts transfer learning through a pre-trained InceptionV3 model at the bottleneck layer, and incorporates a Squeeze and Excitation block in the skip connection, effectively extracting spatial information even from images where 95 % of the background and only 5 % represent the area of interest. A new dataset was developed by gathering oral photographs of dental caries from two hospitals, with advanced augmentation techniques applied. The LCSNet architecture demonstrated an accuracy of 97.36 %, precision of 73.1 %, recall of 70.2 %, an F1-Score of 71.14 %, and an Intersection-over-Union (IoU) of 56.8 %. Expert dentists confirmed that the LCSNet model proposed in this in vivo study accurately segments the position and texture of dental caries. Both qualitative and quantitative performance analyses, along with comparative analyses of efficiency and computational requirements, were conducted with other deep learning models. The proposed model outperforms existing deep learning models and shows significant potential for integration into a smartphone application-based oral disease detection system, potentially replacing some conventional clinically adapted methods.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"12 ","pages":"Article 100304"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145361516","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Cardiovascular disease causes 17.9 million deaths annually, yet current AI systems achieve ∼82 % accuracy without uncertainty quantification—limiting clinical utility where prediction confidence directly guides life-saving treatment decisions.
Objective
We developed an uncertainty-aware hybrid optimization framework for robust CVD detection that provides clinicians with both risk predictions and confidence intervals, enabling personalized decision-making under real-world clinical conditions.
Methods
Our clinical translation framework integrates multiple complementary AI models (Gaussian processes, gradient-boosted trees, Transformers) through uncertainty-guided optimization. Key clinical innovations include: (1) real-time uncertainty calibration responding to data quality variations, (2) dynamic model weighting adapting to individual patient characteristics, and (3) interpretable confidence intervals supporting clinical decision protocols.
Results
Clinical validation on 12,458 CVD patients from MIMIC-III and UK Biobank demonstrated clinically significant improvements: +1.4 % AUC (0.853 vs 0.839, p < 0.01) translating to 50 additional correct diagnoses per 10,000 patients, +1.5 % balanced accuracy, and 20 % better uncertainty calibration. The framework maintained robust performance (>80 % AUC) under realistic clinical noise while providing reliable confidence intervals across all risk levels.
Clinical translation
This framework delivers immediate clinical utility through real-time inference (<2s), FHIR-compliant EHR integration, and physician-validated uncertainty interpretation. Implementation prevents an estimated 50 missed diagnoses and 23 unnecessary procedures per 10,000 patients screened annually.
Conclusions
Our uncertainty-aware framework represents the first clinically ready AI system providing both accurate CVD risk assessment and trustworthy confidence measures, directly addressing physician adoption barriers and supporting personalized cardiovascular care.
背景：心血管疾病每年导致1790万人死亡，但目前的人工智能系统在没有不确定性量化的情况下达到了82%的准确率，这限制了临床实用性，预测置信度直接指导挽救生命的治疗决策。目的：我们开发了一个不确定性感知的混合优化框架，用于稳健的心血管疾病检测，为临床医生提供风险预测和置信区间，从而在现实临床条件下实现个性化决策。方法通过不确定性导向优化，sour临床翻译框架集成了多个互补的人工智能模型（高斯过程、梯度增强树、变形金刚）。关键的临床创新包括：(1)响应数据质量变化的实时不确定度校准，(2)适应个体患者特征的动态模型加权，以及(3)支持临床决策方案的可解释置信区间。结果：来自MIMIC-III和UK Biobank的12,458例CVD患者的临床验证显示出临床显着改善：+ 1.4%的AUC （0.853 vs 0.839, p < 0.01）转化为每10,000例患者额外50例正确诊断，+ 1.5%的平衡准确性和20%的不确定度校准。该框架在真实的临床噪声下保持稳健的性能（80% AUC），同时在所有风险水平上提供可靠的置信区间。临床翻译该框架通过实时推理（<2s）、符合fhir的EHR集成和医生验证的不确定性解释，提供即时的临床效用。每年每1万名接受筛查的患者中，估计有50例漏诊和23例不必要的手术得到预防。结论我们的不确定性感知框架代表了第一个临床就绪的人工智能系统，提供准确的心血管疾病风险评估和可信赖的信心措施，直接解决医生采用障碍并支持个性化心血管护理。
{"title":"Uncertainty-aware hybrid optimization for robust cardiovascular disease detection: A clinical translation framework","authors":"Tamanna Jena , Rahul Suryodai , Desidi Narsimha Reddy , Kambala Vijaya Kumar , Elangovan Muniyandy , N.V. Phani Sai Kumar","doi":"10.1016/j.ibmed.2025.100302","DOIUrl":"10.1016/j.ibmed.2025.100302","url":null,"abstract":"<div><h3>Background</h3><div>Cardiovascular disease causes 17.9 million deaths annually, yet current AI systems achieve ∼82 % accuracy without uncertainty quantification—limiting clinical utility where prediction confidence directly guides life-saving treatment decisions.</div></div><div><h3>Objective</h3><div>We developed an uncertainty-aware hybrid optimization framework for robust CVD detection that provides clinicians with both risk predictions and confidence intervals, enabling personalized decision-making under real-world clinical conditions.</div></div><div><h3>Methods</h3><div>Our clinical translation framework integrates multiple complementary AI models (Gaussian processes, gradient-boosted trees, Transformers) through uncertainty-guided optimization. Key clinical innovations include: (1) real-time uncertainty calibration responding to data quality variations, (2) dynamic model weighting adapting to individual patient characteristics, and (3) interpretable confidence intervals supporting clinical decision protocols.</div></div><div><h3>Results</h3><div>Clinical validation on 12,458 CVD patients from MIMIC-III and UK Biobank demonstrated clinically significant improvements: +1.4 % AUC (0.853 vs 0.839, p < 0.01) translating to 50 additional correct diagnoses per 10,000 patients, +1.5 % balanced accuracy, and 20 % better uncertainty calibration. The framework maintained robust performance (>80 % AUC) under realistic clinical noise while providing reliable confidence intervals across all risk levels.</div></div><div><h3>Clinical translation</h3><div>This framework delivers immediate clinical utility through real-time inference (<2s), FHIR-compliant EHR integration, and physician-validated uncertainty interpretation. Implementation prevents an estimated 50 missed diagnoses and 23 unnecessary procedures per 10,000 patients screened annually.</div></div><div><h3>Conclusions</h3><div>Our uncertainty-aware framework represents the first clinically ready AI system providing both accurate CVD risk assessment and trustworthy confidence measures, directly addressing physician adoption barriers and supporting personalized cardiovascular care.</div></div>","PeriodicalId":73399,"journal":{"name":"Intelligence-based medicine","volume":"12 ","pages":"Article 100302"},"PeriodicalIF":0.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145319571","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号