Skip to main content

Table 2 Detection rate of the CNN model in the testing set, based on case level

From: Convolutional neural network for detecting rib fractures on chest radiographs: a feasibility study

CNN model

Chest radiograph

Total

With rib fractures

Without rib fractures

Detected fractures

151

48

199

Undetected fractures

11

185

196

Total

162

233

395

  1. Sensitivity is 93.2% [TP/(TP + FN) × 100% = 151/162 × 100%]. Specificity is 79.4% [TN/(TN + FP) × 100% = 185/233 × 100%]. The positive predictive value (PPV) is 75.9% [TP/(TP + FP) × 100% = 151/199 × 100%]. The negative predictive value (NPV) is 94.4% [TN/(TN + FN) × 100% = 185/196 × 100%]. Accuracy is 85.1% [(TP + FN)/(TP + FN + TN + FN) × 100% = (151 + 185)/395 × 100%]
  2. CNN convolutional neural network, TP true positive, FN false negative, TN true negative, FP false positive
Back to article page

BMC Medical Imaging

ISSN: 1471-2342

Contact us