Table 4.

Studies with preoperative applications.

Authors (Ref)	Outcome	Data Source	Type of AI	Findings
Nosrati et al. [26]	Intraoperative outcomes	Subsurface cues such as pulsation patterns, textures, and colours within the operative field and preoperative imaging	Machine learning	The framework has application in non- visible and partially occluded structures during PN
DiDio et al. [27]	Intraoperative outcomes	Preoperative CT Pyelography	Computer vision to create HA3D model	The model used to selectively apply pressure to the artery during PN
Klen et al. [28]	Postoperative outcomes	Data from 1099 operated RARC patients	Machine Learning	Identifying patients who are at high risk for complications after RC, and additional factors identified via ASA score, CPD, ACCI, and CHF
Checcucci et al. [29]	Postoperative outcomes	3D and non 3D prostate models from RARP and mpMRI	Multivariable linear regression models	No extracapsular extension in mpMRI and the use of 3D models during RARP lowered the incidence of positive margins

Abbreviations: PN = partial nephrectomy; CT = computed tomography; HA3D = hyper-accuracy 3D model; RC = radical cystectomy; RARC = robotic-assisted radical cystectomy; ASA = American Society of Anesthesiologists; CPD = chronic pulmonary disease; ACCI = age-adjusted Charlson comorbidity index; CHF = chronic heart failure; RARP = robotic-assisted radical prostatectomy; mpMRI = multiparametric MRI.