Lung Imaging

During the COVID-19 pandemic, lung imaging takes a key role in complementing biomolecular testing. The rise of artificial intelligence induced a quantum leap in medical image analysis and AI has proven equipollent to healthcare professionals in several diseases. This rapid development is a double-edged sword: The potential to save time, cost and lives is significant, but AI-accelerated medical imaging must still fully demonstrate its ability in remediating diseases such as COVID-19.

The aim of our work is to assess the adoption of AI technologies in medical imaging for new diseases which is exemplified by the COVID-19 pandemic. We conducted the, to date, largest systematic review of the literature addressing the utility AI in Medical Imaging for COVID-19 management. We provide a comprehensive comparison of the major modalities (CT, X-Ray and ultrasound) by discussing the current clinical evidence and reviewing the state-of-the-art in AI. We identify a disparity between clinicians and the AI community, both in the focus on imaging modalities and performed tasks.

COVID-19 causally evoked a 2-4-fold increase in publications on AI in medical imaging compared to 2019 which motivated us to perform a systematic, manual survey of 463 publications about AI on lung imaging data of COVID-19 patients. Despite conservative regulations from medical bodies and explicit discouragement of using imaging for diagnostic purposes, practice has shown value of medical imaging for patient stratification and, in addition, efforts from the AI community are heavily skewed towards automatizing detection. Moreover, we report a mismatch in the focus of the AI community that payed disproportionally high attention to X-Ray compared to CT. Ultrasound is neglected despite clinical superiority to X-Ray for diagnostic tasks. The rise of lung ultrasound has manifested scientifically throughout the last decade and is corroborated by our market analysis suggesting that ultrasound will expand its global lead of the imaging market. Taken together, this gap is instructive for the community as it identifies acceleration of ultrasound image analysis as a key goal of future research, especially given that it is the only sensitive modality that is widely available globally.

The final part of our contribution derives key challenges for AI acceleration, attempting to pave a road for the community. We touch on generalization and adversarial robustness, discuss the role of interpretable AI and identify economic and logistic factors like on-device analysis of portable machines and perceived barriers of imaging dissemination.

In sum, we hope to provide a coherent and comprehensive summary on the plethora of recent progress of AI on COVID-19 imaging for a clinical and mildly technical target audience.