Published: Sept. 30, 2024
Language: Английский
Computational and Structural Biotechnology Journal, Journal Year: 2024, Volume and Issue: 23, P. 1510 - 1521
Published: April 12, 2024
Fully supervised learning methods necessitate a substantial volume of labelled training instances, process that is typically both labour-intensive and costly. In the realm medical image analysis, this issue further amplified, as annotated images are considerably more scarce than their unlabelled counterparts. Consequently, leveraging to extract meaningful underlying knowledge presents formidable challenge in analysis. This paper introduces simple triple-view unsupervised representation model (SimTrip) combined with architecture loss function, aiming learn inherent efficiently from data small batch size. With extracted data, our demonstrates exemplary performance across two datasets. It achieves using only partial labels outperforms other state-of-the-art methods. The method we present herein offers novel paradigm for learning, establishing baseline poised inspire development intricate SimTrip-based spectrum computer vision applications. Code user guide released at https://github.com/JerryRollingUp/SimTripSystem, system also runs http://43.131.9.159:5000/.
Language: Английский
Citations
19
BMC Medical Informatics and Decision Making, Journal Year: 2024, Volume and Issue: 24(1)
Published: Jan. 8, 2024
Abstract Background Accurate diagnosis and early treatment are essential in the fight against lymphatic cancer. The application of artificial intelligence (AI) field medical imaging shows great potential, but diagnostic accuracy lymphoma is unclear. This study was done to systematically review meta-analyse researches concerning performance AI detecting using for first time. Methods Searches were conducted Medline, Embase, IEEE Cochrane up December 2023. Data extraction assessment included quality independently by two investigators. Studies that reported an model/s detection systemic review. We extracted binary data obtain outcomes interest: sensitivity (SE), specificity (SP), Area Under Curve (AUC). registered with PROSPERO, CRD42022383386. Results Thirty studies systematic review, sixteen which meta-analyzed a pooled 87% (95%CI 83–91%), 94% (92–96%), AUC 97% (95–98%). Satisfactory observed subgroup analyses based on algorithms types (machine learning versus deep learning, whether transfer applied), sample size (≤ 200 or > 200), clinicians models geographical distribution institutions (Asia non-Asia). Conclusions Even if possible overestimation further better standards needed, we suggest may be useful diagnosis.
Language: Английский
Citations
14
Published: May 9, 2024
Language: Английский
Citations
9
Clinical and Experimental Medicine, Journal Year: 2024, Volume and Issue: 24(1)
Published: Aug. 6, 2024
Abstract Traditional manual blood smear diagnosis methods are time-consuming and prone to errors, often relying heavily on the experience of clinical laboratory analysts for accuracy. As breakthroughs in key technologies such as neural networks deep learning continue drive digital transformation medical field, image recognition technology is increasingly being leveraged enhance existing processes. In recent years, advancements computer have led improved efficiency identification cells smears through use technology. This paper provides a comprehensive summary steps involved utilizing algorithms diagnosing diseases smears, with focus malaria leukemia. Furthermore, it offers forward-looking research direction development cell pathological detection system.
Language: Английский
Citations
4
Published: May 9, 2024
Language: Английский
Citations
3
Published: May 9, 2024
Language: Английский
Citations
2
Published: May 9, 2024
Language: Английский
Citations
2
Image and Vision Computing, Journal Year: 2024, Volume and Issue: 151, P. 105298 - 105298
Published: Oct. 11, 2024
Language: Английский
Citations
2
Diagnostics, Journal Year: 2023, Volume and Issue: 13(20), P. 3234 - 3234
Published: Oct. 17, 2023
Brain tumors pose a complex and urgent challenge in medical diagnostics, requiring precise timely classification due to their diverse characteristics potentially life-threatening consequences. While existing deep learning (DL)-based brain tumor (BTC) models have shown significant progress, they encounter limitations like restricted depth, vanishing gradient issues, difficulties capturing intricate features. To address these challenges, this paper proposes an efficient skip connections-based residual network (ESRNet). leveraging the (ResNet) with connections. ESRNet ensures smooth flow during training, mitigating problem. Additionally, architecture includes multiple stages increasing numbers of blocks for improved feature pattern recognition. utilizes from ResNet architecture, featuring connections that enable identity mapping. Through direct addition input tensor convolutional layer output within each block, preserve flow. This mechanism prevents gradients, ensuring effective information propagation across layers training. Furthermore, integrates downsampling techniques stabilizing batch normalization layers, which collectively contribute its robust reliable performance. Extensive experimental results reveal significantly outperforms other approaches terms accuracy, sensitivity, specificity, F-score, Kappa statistics, median values 99.62%, 99.68%, 99.89%, 99.47%, 99.42%, respectively. Moreover, achieved minimum performance metrics, including accuracy (99.34%), sensitivity (99.47%), specificity (99.79%), F-score (99.04%), statistics (99.21%), underscore exceptional effectiveness BTC. Therefore, proposed showcases efficiency BTC, holding potential revolutionize clinical diagnosis treatment planning.
Language: Английский
Citations
6
Published: May 9, 2024
Language: Английский
Citations
1