Hybrid Deep-Learning and Machine-Learning Models for Predicting COVID-19

Table 1

Literature review summary.

Author	Data set size	Image type	Disease type	ML	CNN model	TL models	Date	Max. accuracy
Minaee et al. [21]	5000	X-ray and CT	COVID-19	No	No	ResNet50, ResNet18, DenseNet-121, and SqueezeNet	2020	98%
Jain et al. [22]	6432	X-ray and CT	COVID-19	No	No	Xception	2020	97.97%
Hussain et al. [23]	558	X-ray	COVID-19 and viral pneumonia	Yes	No	No	2020	97.56%
Sekeroglu et al. [6]	6200	X-ray	COVID-19 and viral pneumonia	Yes	Yes	VGG19, MobileNet, inception, Xception, and inception ResNet	2020	99.18%
Linda Wang et al. [24]	13,975	X-ray	COVID-19 and viral pneumonia	No	Yes	VGG-19 and ResNet-50	2020	98.9%
Dingding Wang et al. [25]	1102	X-ray	COVID-19	Yes	No	VGG-16, Xception, ResNet50, and DenseNet121	2020	99.38%
Rahimzadeh et al. [26]	11302	X-ray	COVID-19 and viral pneumonia	No	No	Xception and ResNet50V2	2020	99.50%
Rahul et al. [27]	5840	X-ray	COVID-19 and viral pneumonia	Yes	No	ResNet152	2020	97.7%
Our work	4646	X-ray	COVID-19 and viral pneumonia	Yes	Yes	VGG16, VGG19, ReNet50, and MobileNet	2021	99.82