Introduction
The use of artificial intelligence (AI) algorithms for the detection and diagnosis of medical conditions has been gaining significant momentum. Among the various medical areas, lymphoma, a form of cancer that affects the lymphatic system, is of particular interest. However, to date, a comprehensive systematic review and meta-analysis on the diagnostic accuracy of AI in lymphoma using medical imaging have not been conducted. This study aims to fill this knowledge gap and provide a comprehensive assessment of the diagnostic accuracy of AI algorithms in detecting lymphoma using medical imaging.
Methodology
A systematic search was conducted in medical databases, engineering and technology databases, and clinical trial registries. Articles published in English between January 2000 and August 2020 were selected for inclusion. The selected studies were assessed for quality using an adapted QUADAS-AI assessment tool.
Results
Our results show that AI algorithms can be used for the detection of lymphoma using medical imaging with an SE of 87% and SP of 94%. The pooled results demonstrated an AUC of 97%, aligning closely with the performance of established conventional diagnostic methods for lymphoma. However, significant between-study heterogeneity emerged within the comparison of AI-assisted clinicians and pure clinicians.
Conclusion
In conclusion, AI algorithms exhibit commendable performance in detecting lymphoma. However, the heterogeneity among the included studies makes it difficult to generalize the results with larger sample sizes or in other countries. Therefore, rigorous subgroup analyses and meta-regression for different sample sizes, various algorithms applied, geographical distribution and AI algorithms-assisted clinicians versus pure clinicians should be conducted. Additionally, future non-trivial applications of AI in medical settings may require physicians to combine pieces of demographic information with image data.
This study has several limitations that merit consideration. Firstly, a relatively small number of studies were available for inclusion, which could have skewed diagnostic performance estimates. Additionally, the restricted number of studies addressing diagnostic accuracy in each subgroup, such as specific lymphoma subtypes and medical imaging modalities, prevented a comprehensive assessment of potential sources of heterogeneity. Consequently, the generalizability of our conclusions to diverse lymphoma subtypes and varied medical imaging modalities, particularly without the integration of AI models at this current stage, could be limited. Secondly, we did not conduct a quality assessment for transparency since current diagnostic accuracy reporting standards (STARD-2015) is not fully applicable to the specifics and nuances of AI research. Thirdly, several included studies have methodological deficiencies or are poorly reported, which may need to be interpreted with caution. Furthermore, the wide range of imaging technology, patient populations, pathologies, study designs and AI models used may have affected the estimation of diagnostic accuracy of AI algorithms. Finally, this study only evaluated studies reporting the diagnostic performance of AI using medical image, which is difficult to extend to the impact of AI on patient treatment and outcomes.
Future efforts should focus on robust designs and high-quality reporting to improve the performance and reporting of AI algorithms in detecting lymphoma. These efforts should include the establishment of interconnected networks between medical institutions, fostering unified standards for data acquisition, labeling procedures, and imaging protocols to enable external validation in professional environments, and the encouragement of AI researchers in medical imaging to report studies that do not reject the null hypothesis. Additionally, the development of “customized” AI models tailored to specific domains, such as lymphoma, head and neck cancer, or brain MRI, may be a promising approach.
Priyam Gera