Page 93 - ITPS-7-3

INNOSC Theranostics and
Pharmacological Sciences Medical imaging technology

imaging method can only obtain partial information about disease risk assessment and prediction. By analyzing vast
18
the body and cannot provide a sufficient basis for accurate repositories of image data alongside clinical information,
diagnosis. At present, no imaging results can completely AI facilitates the identification of disease risk factors,
replace clinical examinations, and imaging needs to be enabling proactive interventions and personalized medical
combined with other examination methods to achieve management. Moreover, AI-driven image reconstruction
accurate diagnosis. and enhancement methodologies refine image quality and
In recent years, advancements in medical imaging resolution, empowering physicians with clearer and more
technology have significantly enhanced its role in clinical informative visuals for precise diagnostic and therapeutic
diagnosis. However, several challenges remain, such decision-making. 19,20 In essence, the fusion of medical
as avoiding interference from various factors during imaging technology and AI signifies a paradigm shift in
the imaging process and improving imaging quality. medical practice, underpinning enhanced diagnostic
In multi-modal imaging, there is a need for better accuracy, personalized care, and elevated standards of
integration of images from different sources after certain health-care delivery. 21
transformations to obtain comprehensive information 6. Conclusion
about the anatomy and function of the body from a single
image. Some imaging methods still require improvement. The development of medical imaging technology will
For instance, MRI primarily relies on hydrogen molecular accelerate, with applications becoming more mature, image
imaging, which poses challenges for imaging areas like quality becoming clearer, and the advantages of imaging
the alveoli, which are mainly gas filled and contain increasingly integrated. This progress brings new hope to
almost no water. From a safety perspective, although countless patients and will contribute significantly to the
the concentration of radioactive tracers used in nuclear prevention, early diagnosis, and treatment of diseases.
medicine imaging is very low, these radioactive tracers
cause radiation exposure in the patient’s body until they Acknowledgments
are eliminated or decay. Therefore, when selecting a None.
radioactive tracer, its half-life and radiation dose should
be considered. 14 Funding
In medical image analysis, convolutional neural None.
networks (CNNs) achieve over 90% accuracy, aiding in
lesion identification. They efficiently extract features, Conflict of interest
15
expediting diagnosis and reducing physician workload, The authors declare that they have no competing interests.
thereby improving diagnostic consistency. CNNs find
broad applications, revolutionizing medical imaging Author contributions
16
across domains such as radiology and pathology. Conceptualization: Bin Zhang, Chin Siang Kue
CNNs distinguish themselves from other artificial neural Writing – original draft: Bin Zhang
networks primarily through their incorporation of Writing – review & editing: All authors
convolutional layers. This addition significantly enhances
the performance of neural networks, spurring the creation Ethics approval and consent to participate
of numerous convolutional models and techniques
aimed at further optimization and innovation. A novel Not applicable.
15
approach, termed the differential CNN, coupled with Consent for publication
simultaneous multidimensional filter implementation,
has enhanced the performance of CNNs while mitigating Not applicable.
the computational cost associated with conventional
methods. Techniques such as deep learning and machine Availability of data
17
learning algorithms, including CNNs, have demonstrated Not applicable.
remarkable success in tasks such as tumor detection,
lesion localization, and disease classification, substantially References
augmenting the precision and effectiveness of image 1. Abhisheka B, Biswas SK, Purkayastha B, Das D, Escargueil A.
interpretation. Furthermore, AI leverages extensive data Recent trend in medical imaging modalities and their
analysis and pattern recognition capabilities to unveil applications in disease diagnosis: A review. Multimed Tools
intricate patterns and features within images, aiding in Appl. 2024;83(14):43035-43070.

Volume 7 Issue 3 (2024) 9 doi: 10.36922/itps.3360