Light Weighted Healthcare CNN Model to Detect Prostate Cancer on Multiparametric MRI

Table 1

Quantitative comparison of the results of different network models for prostate cancer lesion segmentation.

Network model	MRI type	DSC	Sensitivity	Specificity	PPV
U Network model	ADC	0.621	0.685	0.987	0.456
V Network model		0.678	0.859	0.962	0.562
RUNM		0.654	0.802	0.945	0.552
MRCNN		0.652	0.625	0.985	0.621
U Network model	T2w	0.421	0.785	0.845	0.265
V Network model		0.386	0.526	0.823	0.285
RUNM		0.365	0.662	0.956	0.296
MRCNN		0.456	0.456	0.945	0.562
U Network model	Hybrid	0.538	0.856	0.965	0.548
V Network model		0.569	0.852	0.985	0.456
RUNM		0.621	0.785	0.9642	0.526
MRCNN		0.785	0.741	0.947	0.685