Enhancing Clinical Diagnosis With Convolutional Neural Networks: Developing High-Accuracy Deep Learning Models for Differentiating Thoracic Pathologies DOI Open Access

Kartik K. Goswami,

Nathaniel Tak,

Arnav Wadhawan

и другие.

Cureus, Год журнала: 2024, Номер unknown

Опубликована: Июль 26, 2024

Background The use of computational technology in medicine has allowed for an increase the accuracy clinical diagnosis, reducing errors through additional layers oversight. Artificial intelligence technologies present potential to further augment and expedite accuracy, quality, efficiency at which diagnosis can be made when used as adjunctive tool. Such techniques, if found accurate reliable their diagnostic acuity, implemented foster better decision-making, improving patient quality care while healthcare costs. Methodology This study convolution neural networks develop a deep learning model capable differentiating normal chest X-rays from those indicating pneumonia, tuberculosis, cardiomegaly, COVID-19. There were 3,063 X-rays, 3,098 pneumonia 2,920 COVID-19 2,214 554 tuberculosis Kaggle that training validation. was trained recognize patterns within efficiently these diseases patients treated on time. Results results indicated success rate 98.34% incorrect detections, exemplifying high degree accuracy. are limitations this study. Training models require hundreds thousands samples, due variability image scanning equipment techniques images sourced, could have learned interpret external noise unintended details adversely impact Conclusions Further studies implement more universal database-sourced with similar assess diverse but related medical conditions, utilization repeat trials help reliability model. These highlight machine algorithms disease detection X-rays.

Язык: Английский

A novel integrated logistic regression model enhanced with recursive feature elimination and explainable artificial intelligence for dementia prediction DOI Creative Commons
Rasel Ahmed, Nafiz Fahad, M. Saef Ullah Miah

и другие.

Healthcare Analytics, Год журнала: 2024, Номер unknown, С. 100362 - 100362

Опубликована: Сен. 1, 2024

Язык: Английский

Процитировано

6
Emerging intelligent wearable devices for cardiovascular health monitoring DOI
Yiqian Wang, Yang Zou, Zhou Li

и другие.

Nano Today, Год журнала: 2024, Номер 59, С. 102544 - 102544

Опубликована: Ноя. 8, 2024

Язык: Английский

Процитировано

4
Digital biomarkers: Redefining clinical outcomes and the concept of meaningful change DOI Creative Commons
M. Florencia Iulita, Emmanuel Streel, John Harrison

и другие.

Alzheimer s & Dementia Translational Research & Clinical Interventions, Год журнала: 2025, Номер 11(2)

Опубликована: Апрель 1, 2025

MCID (minimal clinically important difference) is a patient-centered concept used in clinical research that represents the smallest change someone living with Alzheimer's disease would identify as important. There are several challenges associated universal application of this construct. progresses differently for each individual, complicating establishment standard accounts individual-level issues. also gradual and evolving disorder, what perceived meaningful can vary significantly at early late stages. People caregivers may have differing perspectives on benefits treatment outcomes, making it more challenging to establish an appropriate MCID. Moreover, trials rely variety tests evaluate cognitive functional impairments. However, these often lack sensitivity early-stage changes affected by variability rater rankings. Digital biomarkers advanced health technologies emerged hot topic modern medicine. They offer promising approach detecting real-time, objective differences improving patient outcomes enabling continuous monitoring, individualized assessments, leveraging artificial intelligence learning complex analytical predictions. while advancements hold great potential, they raise considerations around standardization, accuracy, integration into current frameworks. As new introduced alongside regulatory frameworks, primary focus must remain truly matter people their caregivers, ensuring principle meaningfulness not lost. Minimal difference (MCID) patient's condition be considered meaningful, but defining due its heterogeneity.The perception differ individual level, different stages within same between caregiver.Traditional endpoints detect subtle limited range restrictions, them less effective accurately capturing efficacy.Digital (AI)-driven potential enhance detection providing continuous, monitoring analytics assessments.Both United States Food Drug Administration (FDA) European Medicines Agency (EMA) playing pivotal roles advancing use digital technologies, facilitating evolution frameworks ensure innovations effectively integrated practice.

Язык: Английский

Процитировано

0
Enhancing Clinical Diagnosis With Convolutional Neural Networks: Developing High-Accuracy Deep Learning Models for Differentiating Thoracic Pathologies DOI Open Access

Kartik K. Goswami,

Nathaniel Tak,

Arnav Wadhawan

и другие.

Cureus, Год журнала: 2024, Номер unknown

Опубликована: Июль 26, 2024

Background The use of computational technology in medicine has allowed for an increase the accuracy clinical diagnosis, reducing errors through additional layers oversight. Artificial intelligence technologies present potential to further augment and expedite accuracy, quality, efficiency at which diagnosis can be made when used as adjunctive tool. Such techniques, if found accurate reliable their diagnostic acuity, implemented foster better decision-making, improving patient quality care while healthcare costs. Methodology This study convolution neural networks develop a deep learning model capable differentiating normal chest X-rays from those indicating pneumonia, tuberculosis, cardiomegaly, COVID-19. There were 3,063 X-rays, 3,098 pneumonia 2,920 COVID-19 2,214 554 tuberculosis Kaggle that training validation. was trained recognize patterns within efficiently these diseases patients treated on time. Results results indicated success rate 98.34% incorrect detections, exemplifying high degree accuracy. are limitations this study. Training models require hundreds thousands samples, due variability image scanning equipment techniques images sourced, could have learned interpret external noise unintended details adversely impact Conclusions Further studies implement more universal database-sourced with similar assess diverse but related medical conditions, utilization repeat trials help reliability model. These highlight machine algorithms disease detection X-rays.

Язык: Английский

Процитировано

0