Scientific Reports

Table 2 Case comparison between the proposed model and human experts.

From: Classifying central serous chorioretinopathy subtypes with a deep neural network using optical coherence tomography images: a cross-sectional study


Model	Non-resolving	Non-resolving	Non-resolving	Acute	Chronic	Acute
R1^a	Acute	Acute	Non-resolving	Non-resolving	Acute	Non-resolving
R2^a	Acute	Acute	Non-resolving	Acute	Non-resolving	Non-resolving
R3^a	Acute	Non-resolving	Acute	Acute	Non-resolving	Acute
F1^b	Acute	Acute	Acute	Acute	Acute	Acute
F2^b	Acute	Acute	Acute	Acute	Acute	Acute
F3^b	Acute	Acute	Acute	Acute	Acute	Acute
RS1^c	Acute	Acute	Acute	Non-resolving	Non-resolving	Non-resolving
RS2^c	Acute	Acute	Acute	Non-resolving	Non-resolving	Non-resolving
GT^d	Acute	Acute	Acute	Non-resolving	Non-resolving	Non-resolving

The six images denote the false detections by our model while classifying images into central serous chorioretinopathy subtypes.
^aR1, R2, and R3 denote ophthalmology residents with < 1, 3, and 4 years of experience, respectively.
^bF1, F2, and F3 denote retina fellows with < 1, 2, and 2 years of experience, respectively.
^cRS1 and RS2 refer to retina specialists with > 10 years of experience.
^dGT denotes the ground truth.

Back to article page

Search

Advanced search

Quick links