Diagnosis of Lumbar Spondylolisthesis Using Optimized Pretrained CNN Models

Table 1

Summary of literature review.

Source	Purpose	Major findings	Accuracy (%)
Varcin et al. [23]	Diagnosis of lumbar spondylolisthesis	AlexNet and GoogLeNet were used for spondylolisthesis diagnosis. The model is not suitable in terms of accuracy.	AlexNet: 93.87
			GoogLeNet: 91.67
CoCoci et al. [24]	Pneumonia detection on chest X-ray	MobileNetV3, ShuffleNetV2, and SlimNet models with Android implementations in TensorFlow Lite are presented for constructing intelligent medical devices.	MobileNetV3: 95.9
			ShuffleNetV2: 96.67
			SlimNet: 96.83
Cococi et al. [25]	Disease detection from chest X-ray	A sophisticated medical device is built with an Android and Raspberry Pi-based strategy.	91.22
Basantwani et al. [26]	COVID-19 detection from chest X-rays and CT scans	Android app was built to convert the final model into a TFLite model which could be used in making the Android model	94
Verma et al. [27]	Detecting COVID-19 from chest CT scans	Model’s size is reduced by utilising TensorFlow lite, and model is tuned for speed and latency on edge devices.	99.58
Bushra et al. [28]	Detection of COVID-19 from X-ray images	An Android application is developed which uses the TFLite model	98.65
Zebin and Rezvy [29]	Detection of COVID-19 using chest X-ray	Multiple pretrained models were used for detection of chest disease from X-ray images.	VGG16: 90
			ResNet50: 94.3
			EfficientNetB0: 96.8
Sharma et al. [30]	Multilabel classification of retinal disorders using OCT	Deep-learning-based detection method for screening people with blinding retinal diseases is proposed which can be remedied if detected early.	99.38