RT Journal Article
SR Electronic
T1 Explainable Deep Learning Improves Physician Interpretation of Myocardial Perfusion Imaging
JF Journal of Nuclear Medicine
JO J Nucl Med
FD Society of Nuclear Medicine
SP jnumed.121.263686
DO 10.2967/jnumed.121.263686
A1 Miller, Robert J.H.
A1 Kuronuma, Keiichiro
A1 Singh, Ananya
A1 Otaki, Yuka
A1 Hayes, Sean
A1 Chareonthaitawee, Panithaya
A1 Kavanagh, Paul
A1 Parekh, Tejas
A1 Tamarappoo, Balaji K
A1 Sharir, Tali
A1 Einstein, Andrew J
A1 Fish, Mathews B
A1 Ruddy, Terrence D
A1 Kaufmann, Philipp A.
A1 Sinusas, Albert J
A1 Miller, Edward J
A1 Bateman, Timothy
A1 Dorbala, Sharmila
A1 Di Carli, Marcelo F
A1 Cadet, Sebastien
A1 Liang, Joanna X
A1 Dey, Damini
A1 Berman, Daniel S.
A1 Slomka, Piotr J.
YR 2022
UL http://jnm.snmjournals.org/content/early/2022/05/05/jnumed.121.263686.abstract
AB Rationale: Artificial intelligence may improve accuracy of myocardial perfusion imaging (MPI) but will likely be implemented as an aid to physician interpretation rather than an autonomous tool. Deep learning (DL) has high standalone diagnostic accuracy for obstructive coronary artery disease (CAD), but its influence on physician interpretation is unknown. We assessed whether access to explainable DL predictions improves physician interpretation of MPI. Methods: We selected a representative cohort of patients who underwent MPI with reference invasive coronary angiography. Obstructive CAD, defined as stenosis ≥ 50% in the left main artery or ≥70% in other coronary segments, was present in half of patients. We utilized an explainable DL model (CAD-DL), which was previously developed in a separate population from different sites. Three physicians interpreted studies first with clinical history, stress, and quantitative perfusion, then with all the data plus the DL results. Diagnostic accuracy was assessed using area under the receiver-operating characteristic curve (AUC). Results: In total, 240 patients were included with median age 65 (IQR 58 – 73). The diagnostic accuracy of physician interpretation with CAD-DL (AUC 0.779) was significantly higher compared to physician interpretation without CAD-DL (AUC 0.747, P = 0.003) and stress total perfusion deficit (AUC 0.718, p<0.001). With matched specificity, CAD-DL had higher sensitivity when operating autonomously compared to readers without DL results (p<0.001), but not compared to readers interpreting with DL results (P = 0.122). All readers had numerically higher accuracy with the use of CAD-DL, with AUC improvement 0.02 to 0.05, and interpretation with DL resulted in overall net reclassification improvement of 17.5% (95% CI 9.8% – 24.7%, p<0.001). Conclusion: Explainable DL predictions lead to meaningful improvements in physician interpretation; however, the improvement varied across the readers reflecting the acceptance of this new technology. This technique could be implemented as an aid to physician diagnosis, improving the diagnostic accuracy of MPI.