Deep learning for semi-automated unidirectional measurement of lung tumor size in CT

BACKGROUND

Performing Response Evaluation Criteria in Solid Tumor (RECISTS) measurement is a non-trivial task requiring much expertise and time. A deep learning-based algorithm has the potential to assist with rapid and consistent lesion measurement.

PURPOSE

The aim of this study is to develop and evaluate deep learning (DL) algorithm for semi-automated unidirectional CT measurement of lung lesions.

METHODS

This retrospective study included 1617 lung CT images from 8 publicly open datasets. A convolutional neural network was trained using 1373 training and validation images annotated by two radiologists. Performance of the DL algorithm was evaluated 244 test images annotated by one radiologist. DL algorithm's measurement consistency with human radiologist was evaluated using Intraclass Correlation Coefficient (ICC) and Bland-Altman plotting. Bonferroni's method was used to analyze difference in their diagnostic behavior, attributed by tumor characteristics. Statistical significance was set at p < 0.05.

RESULTS

The DL algorithm yielded ICC score of 0.959 with human radiologist. Bland-Altman plotting suggested 240 (98.4 %) measurements realized within the upper and lower limits of agreement (LOA). Some measurements outside the LOA revealed difference in clinical reasoning between DL algorithm and human radiologist. Overall, the algorithm marginally overestimated the size of lesion by 2.97 % compared to human radiologists. Further investigation indicated tumor characteristics may be associated with the DL algorithm's diagnostic behavior of over or underestimating the lesion size compared to human radiologist.

CONCLUSIONS

The DL algorithm for unidirectional measurement of lung tumor size demonstrated excellent agreement with human radiologist.

2016 New Horizons Lecture: Beyond Imaging-Radiology of Tomorrow

This article is based on the New Horizons lecture delivered at the 2016 Radiological Society of North America Annual Meeting. It addresses looming changes for radiology, many of which stem from the disruptive effects of the Fourth Industrial Revolution. This is an emerging era of unprecedented rapid innovation marked by the integration of diverse disciplines and technologies, including data science, machine learning, and artificial intelligence-technologies that narrow the gap between man and machine. Technologic advances and the convergence of life sciences, physical sciences, and bioengineering are creating extraordinary opportunities in diagnostic radiology, image-guided therapy, targeted radionuclide therapy, and radiology informatics, including radiologic image analysis. This article uses the example of oncology to make the case that, if members in the field of radiology continue to be innovative and continuously reinvent themselves, radiology can play an ever-increasing role in both precision medicine and value-driven health care. © RSNA, 2018.