. 2024 Jan 7;13(2):336. doi: 10.3390/jcm13020336

Table 1.

Summary of the main fields of application of radiomic analyses in PitNETs, with specifically evaluated endpoints and examples of studies included in this review. Abbreviations: CE-T1WI, contrast-enhanced T1-weighted imaging; CS, cavernous sinus; DA, dopamine agonist; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging; MRI, magnetic resonance imaging; N, number; NCA, null-cell adenoma; NFPA, non-functioning pituitary adenoma; PitNET, pituitary neuroendocrine tumor; RT, radiotherapy; SCA, silent corticotroph adenoma; SRL, somatostatin receptor ligand; SS, sphenoid sinus; T1WI, T1-weighted imaging; T2WI, T2-weighted imaging.

General Field of Application	Specific Endpoint	Studies (First Author, Year)	N. of Patients	PitNET Subtypes	Evaluated MRI Sequences	Software Used for Feature Extraction
Prediction of tumor consistency	Distinction between soft and fibrous tumors	Zeynalova, 2019 [6]	55	Any	T2WI	PyRadiomics
		Cuocolo, 2020 [12]	89	Any	T2WI	PyRadiomics
		Wan, 2022 [11]	156	Any	T1WI, T2WI, CE-T1WI	MatLab
		Fan, 2019 [16]	188	Acromegaly	T1WI, T2WI, CE-T1WI	PyRadiomics
Prediction of tumor invasiveness	Prediction of CS invasion	Niu, 2019 [19]	194	Any	T2WI, CE-T1WI	MatLab
	Prediction of CS invasion	Zhang, 2022 [20]	196	Any	CE-T1WI	PyRadiomics
	Prediction of CS, SS, or dura mater invasion	Liu, 2020 [21]	50	Any	DCE-MRI	Omni-Kinetics
Prediction of hormonal secretion patterns	Distinction among hormone secretion profiles	Baysal, 2022 [22]	130	Any	T2WI	PyRadiomics
Prediction of histopathological features	Distinction among Tpit, Pit-1, and SF-1 subfamilies	Peng, 2020 [23]	235	Any	T1WI, T2WI, CE-T1WI	PyRadiomics
	Distinction between SCAs and other NFPA subtypes	Rui, 2022 [24]	302	NFPAs	T1WI, T2WI, CE-T1WI	PyRadiomics
	Distinction between SCAs and other NFPA subtypes	Wang, 2023 [25]	295	NFPAs	T1WI, T2WI, CE-T1WI	PyRadiomics
	Distinction between NCAs and other NFPA subtypes	Zhang, 2018 [26]	112	NFPAs	T1WI, CE-T1WI	MatLab
	Distinction between densely and sparsely granulated adenomas	Park, 2020 [27]	69	Acromegaly	T2WI	PyRadiomics
		Liu, 2021 [28]	49	Acromegaly	T1WI, T2WI, CE-T1WI	PyRadiomics
	Prediction of a high proliferative index (Ki67 ≥ 3%) ^a	Ugga, 2019 [29]	89	Any	T2WI	PyRadiomics
		Li, 2023 [30]	1214	Any	T1WI, T2WI, CE-T1WI	PyRadiomics
		Wang, 2023 [31]	246	Any	CE-T1WI	LIFEx
		Fan, 2020 [32]	138	Acromegaly	T1WI, T2WI, CE-T1WI	PyRadiomics
Prediction of response to surgical treatment	Prediction of post-surgical recurrence or regrowth	Galm, 2018 [34]	78	NFPAs	T1WI	ImageJ
		Machado, 2020 [35]	27	NFPAs	CE-T1WI	PyRadiomics
		Zhang, 2020 [36]	50	NFPAs	T2WI, CE-T1WI	Python
		Zhang, 2022 [37]	168	Any	CE-T1WI	PyRadiomics (3D-Slicer extension)
	Prediction of post-surgical biochemical remission	Fan, 2019 [38]	163	Functional adenomas	T1WI, T2WI, CE-T1WI	MatLab
Prediction of response to non-surgical therapies	Response to first-generation SRLs	Kocak, 2019 [44]	47	Acromegaly	T2WI	PyRadiomics
	Response to first-generation SRLs	Galm, 2020 [45]	64	Acromegaly	T1WI	ImageJ
	Response to DAs	Park, 2021 [46]	177	Prolactinomas	T2WI	PyRadiomics
	Response to RT	Fan, 2019 [47]	57	Acromegaly	T1WI, T2WI, CE-T1WI	PyRadiomics

^a In the study by Wang et al. (2023) [31], the definition of a high proliferation/aggressive behavior was based on the presence of at least two of the following three characteristics: Ki67 ≥ 3%, high mitotic count (≥2/10 high power fields), or positive staining for p53.