Table 4.

Table representing the performances of various State-of-art models.

Approach	Dataset	Sensitivity	Specificity	Accuracy	F1-Score
SW-FFANN [56]	PIMA	0.85	0.96	0.91	N/A
ResNet18 [55]	PIMA	0.88	0.66	0.80	0.85
ResNet50 [55]	PIMA	0.94	0.57	0.80	0.85
SVM [55]	PIMA	0.95	0.83	0.90	0.93
SVM + RF + MLP + ANFIS [57]	Real-time sensor data	N/A	N/A	0.90	N/A
Deep transfer learning [58]	OhioT1DM dataset	0.59	0.98	0.95	0.61
LR [59]	PPG Signal data	0.73	0.64	0.69	N/A
GSVM [60]	PPG Signal data	0.79	0.83	0.81	N/A
Bayesian classifier [61]	PPG Signal data	1.00	0.87	0.93	N/A
SVM [62]	PPG Signal data	0.98	0.96	0.97	N/A
AdaBoost [63]	ECG Signal data	0.92	0.88	0.90	N/A
IGRNet [64]	ECG Signal data	0.80	0.77	0.77	N/A
GoogLeNet [64]	ECG Signal data	0.69	0.83	0.75	N/A
AlexNet [64]	ECG Signal data	0.77	0.82	0.74	N/A
HoG with K-NN [64]	ECG Images	0.70	0.79	0.76	N/A
AlexNet [19]	Spectrogram images	0.93	0.96	0.95	0.92
ResNet [19]	Spectrogram images	0.93	0.96	0.95	0.92
KNN [12]	Spectrogram images	0.89	0.89	0.86	0.86
SVM [12]	Spectrogram images	0.91	0.91	0.89	0.89
DenseNet-121 [65]	Spectrogram images	0.72	0.73	0.98	N/A
ResNet with SVM [66]	Spectrogram images	0.98	0.98	0.98	0.98
Proposed Approach	Spectrogram images	0.97	0.96	0.96	0.93