Abstract
Carbohydrate counting is essential for well-controlled blood glucose in people with type 1 diabetes, but to perform it precisely is challenging, especially for Thai foods. Consequently, we developed a deep learning-based system for automatic carbohydrate counting using Thai food images taken from smartphones. The newly constructed Thai food image dataset contained 256,178 ingredient objects with measured weight for 175 food categories among 75,232 images. These were used to train object detector and weight estimator algorithms. After training, the system had a Top-1 accuracy of 80.9% and a root mean square error (RMSE) for carbohydrate estimation of <10 g in the test dataset. Another set of 20 images, which contained 48 food items in total, was used to compare the accuracy of carbohydrate estimations between measured weight, system estimation, and eight experienced registered dietitians (RDs). System estimation error was 4%, while estimation errors from nearest, lowest, and highest carbohydrate among RDs were 0.7, 25.5, and 7.6%, respectively. The RMSE for carbohydrate estimations of the system and the lowest RD were 9.4 and 10.2, respectively. The system could perform with an estimation error of <10 g for 13/20 images, which placed it third behind only two of the best performing RDs: RD1 (15/20 images) and RD5 (14/20 images). Hence, the system was satisfactory in terms of accurately estimating carbohydrate content, with results being comparable with those of experienced dietitians.
Highlights
In 2019, type 1 diabetes (T1D) was estimated to affect 1,110,100 persons aged
The system consisted of three independent algorithms, namely, the convolutional neural network (CNN)-based object detector, segmentation unit, and neural network regressionbased weight estimation unit
The difference between the means of the calculated carbohydrate content of Measured (30.3 ± 14.8 g) and Estimated (29.2 ± 15.3) was small enough that the paired t-test could not detect the difference as indicated by a p-value of 0.625, whereas the paired t-test was able to detect the difference between the highest (36.7 ± 19.7) of RD4 and the lowest (25.4 ± 16.5) of RD5 with p-values of 0.047 and 0.03, respectively
Summary
In 2019, type 1 diabetes (T1D) was estimated to affect 1,110,100 persons aged
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have