Deep learning-assisted ultra-accurate smartphone testing of paper-based colorimetric ELISA assays

Article Chemistry, Analytical

Low-Cost, User-Friendly, All-Integrated Smartphone-Based Microplate Reader for Optical-Based Biological and Chemical Analyses

Jose Francisco Bergua et al.

Summary: The quantitative detection of different molecular targets is crucial for various activities. This study presents a smartphone-based device that enables high-precision laboratory analysis directly at the point of care.

ANALYTICAL CHEMISTRY (2022)

Add to Collection

Article Biochemical Research Methods

Rapid segmentation and sensitive analysis of CRP with paper-based microfluidic device using machine learning

Qihong Ning et al.

Summary: This paper presents the fabrication of multi-layer mu PADs for colorimetric detection of CRP using the imprinting method. The detection performance of mu PADs is improved through simulating different lighting conditions and shooting angles, and a machine learning algorithm is used for analysis. The results show that the YOLO model trained in this study can accurately identify all reaction areas, and the residual network algorithm achieves the highest accuracy in the classification task.

ANALYTICAL AND BIOANALYTICAL CHEMISTRY (2022)

Add to Collection

Article Multidisciplinary Sciences

On evaluation metrics for medical applications of artificial intelligence

Steven A. Hicks et al.

Summary: This paper examines the application of machine learning in gastroenterology and provides explanations and interpretations of different metrics. Additionally, an open source tool is developed to assist researchers and clinicians in calculating relevant metrics.

SCIENTIFIC REPORTS (2022)

Add to Collection

Article Engineering, Biomedical

Platelet Detection Based on Improved YOLO_v3

Renting Liu et al.

Summary: Platelet detection and counting are of great significance in the medical field. Traditional methods are time-consuming and require professional doctors. In recent years, deep learning-based methods have provided new possibilities for platelet detection, but due to the difficulty in obtaining datasets and the small size of platelets, there is limited research in this area. This paper conducted experiments using different object detection models and found that YOLO_v3 performs better in accurately detecting platelets. Three improvement ideas were also proposed to further enhance the detection accuracy.

CYBORG AND BIONIC SYSTEMS (2022)

Add to Collection

Article Engineering, Biomedical

Auto-CSC: A Transfer Learning Based Automatic Cell Segmentation and Count Framework

Guangdong Zhan et al.

Summary: Cell segmentation and counting are crucial in the medical field, and convolution neural networks have shown promising results in this area. However, the traditional data-driven approach requires a large number of annotations and can be time-consuming and prone to human error. This paper proposes a novel method that eliminates the need for extensive manual annotations by generating cell image labels using traditional algorithms. The proposed method achieves comparable segmentation and count performance to models trained with a large amount of annotated mixed cell images.

CYBORG AND BIONIC SYSTEMS (2022)

Add to Collection

Article Computer Science, Artificial Intelligence

Comparative analysis of image classification algorithms based on traditional machine learning and deep learning

Pin Wang et al.

Summary: This paper compares and analyzes traditional machine learning and deep learning image classification algorithms, finding that traditional machine learning performs better on small sample datasets, while deep learning has higher recognition accuracy on large sample datasets.

PATTERN RECOGNITION LETTERS (2021)

Add to Collection

Article Chemistry, Analytical

Machine learning-based colorimetric determination of glucose in artificial saliva with different reagents using a smartphone coupled μPAD

Oyku Berfin Mercan et al.

Summary: A portable platform was developed integrating mu PADs with a smartphone application for quantifying glucose concentration in artificial saliva. Three different detection mixtures were used, with the mixture containing TMB showing the highest classification accuracy and inter-phone repeatability under versatile illumination.

SENSORS AND ACTUATORS B-CHEMICAL (2021)

Add to Collection

Article Chemistry, Multidisciplinary

Ultrasensitive and Highly Specific Lateral Flow Assays for Point-of-Care Diagnosis

Yilin Liu et al.

Summary: LFAs, paper-based point-of-care diagnostic tools, have traditionally lower sensitivities and specificities than standard laboratory tests. Recent efforts have focused on optimizing assay kinetics, signal amplification, and sample preamplification to increase sensitivities, as well as improving specificity through high-affinity molecules and assay optimization. With continuing improvements, LFAs may soon offer performance competitive with laboratory techniques while retaining a rapid format.

ACS NANO (2021)

Add to Collection

Article Chemistry, Multidisciplinary

Smartphone-Enabled Paper-Based Hemoglobin Sensor for Extreme Point-of-Care Diagnostics

Sujay K. Biswas et al.

Summary: The study introduces a paper-based sensor integrated with a smartphone application for rapid and affordable detection of hemoglobin concentration, showing high sensitivity, specificity, and accuracy in both laboratory and field settings, particularly beneficial for underserved communities.

ACS SENSORS (2021)

Add to Collection

Review Biochemical Research Methods

Smartphone-based imaging systems for medical applications: a critical review

Brady Hunt et al.

Summary: Smartphones have a wide range of applications in healthcare, especially with specialized attachments. Research on smartphone-based imaging systems and evaluation of current studies provide guidelines for improving research impact.

JOURNAL OF BIOMEDICAL OPTICS (2021)

Add to Collection

Review Chemistry, Multidisciplinary

Microfluidic Paper-Based Analytical Devices: From Design to Applications

Eka Noviana et al.

Summary: Microfluidic paper-based analytical devices (muPADs) have gained significant interest as a promising analytical platform in the past decade, offering unique advantages over traditional microfluidics. This comprehensive Review highlights fabrication methods, device designs, and detection strategies, as well as the growing applications of muPADs, while discussing the field's need for further advancement to realize its full potential.

CHEMICAL REVIEWS (2021)

Add to Collection

Article Biochemical Research Methods

Development of a smartphone-based lateral-flow imaging system using machine-learning classifiers for detection of Salmonella spp.

Hyun Jung Min et al.

Summary: This study introduced a smartphone-based lateral-flow assay analyzer for detecting Salmonella spp. using machine-learning algorithms, achieving high accuracy and providing a new method for distinguishing ambiguous concentrations of test lines.

JOURNAL OF MICROBIOLOGICAL METHODS (2021)

Add to Collection

Article Engineering, Electrical & Electronic

Smartphone-based DNA diagnostics for malaria detection using deep learning for local decision support and blockchain technology for security

Xin Guo et al.

Summary: A smartphone-based end-to-end platform for multiplexed DNA diagnosis of malaria using paper-based microfluidic diagnostic tests, deep learning algorithms, and blockchain technology was validated through field tests in rural Uganda, achieving over 98% accuracy in identifying tested cases. The platform also offers secure geotagged diagnostic information for potential integration into infectious disease surveillance frameworks.

NATURE ELECTRONICS (2021)

Add to Collection

Article Chemistry, Analytical

Smartphone-based colorimetric detection system for portable health tracking

Samira Balbach et al.

Summary: Colorimetric tests for at-home health monitoring have become popular due to their reduced costs and ease of operation, but developing digital systems to interface these sensors remains a challenge. Efforts have been made to develop portable optical readout systems, such as smartphones, but their use in daily settings is limited by optical noise and low sensitivity.

ANALYTICAL METHODS (2021)

Add to Collection

Article Computer Science, Information Systems

Performance Evaluation of Deep Learning Classification Network for Image Features

Qiang Li et al.

Summary: Researchers constructed different datasets to test the performance of nine mainstream image classification networks, and found that some of these networks showed better classification performance. The experimental results analyzed the sensitivity of factors that influence the stability of image classification networks.

IEEE ACCESS (2021)

Add to Collection

Article Chemistry, Analytical