Value of Intra-Perinodular Textural Transition Features from MRI in Distinguishing Between Benign and Malignant Testicular Lesions

Figures & data

Figure 1 Flowchart for patient selection.

Table 1 MRI Protocol for Testicular Examination

Sequence Parameters	T1WI	T2WI
Imaging plane	Axial	Axial, sagittal, coronal
Repetition time (ms)	750	6500–6870
Echo time (ms)	13	104
Slice thickness (mm)	3~5	3~5
Slice gap (mm)	0–0.5	0–0.5
Field of view (mm^2)	300 × 300	180 × 180
Matrix	320 × 240	384 × 320
Flip angle (degree)	160	160
Average	1	2

Figure 2 Flowchart of feature extraction. (A) All patients underwent preoperative magnetic resonance imaging. (B) Testicular lesions were delineated by stacking regions of interest (ROI) slice-by-slice on the transverse T2-weighted images. (C) The Ipris and histogram features were extracted from T2WI by Python software and IF software, respectively.

Figure 3 The distribution of intra- and interobserver intraclass correlation coefficients (ICCs) values for all the Ipris and histogram features (A and B).

Table 2 Diagnostic Performance of the 12 Selected Features

Features	p-value	AUC (95% CI)	Sensitivity	Specificity
			(%)	(%)
Histogram_TotalEnergy	<0.001	0.808	50.9	96
		(0.705–0.888)
Histogram_Energy	<0.001	0.807	74.6	72
		(0.704–0.887)
Histogram_Range	0.021	0.654	60	68
		(0.540–0.757)
Ipris_shell1_id_std	0.001	0.708	65.5	76
		(0.595–0.804)
Ipris_shell2_id_mean	0.002	0.69	43.6	96
		(0.577–0.789)
Ipris_shell1_ge_max	0.022	0.651	54.6	72
		(0.536–0.754)
Ipris_shell1_gd_min	0.036	0.649	58.2	72
		(0.534–0.752)
Ipris_shell1_gs_max	0.024	0.648	54.6	72
		(0.533–0.751)
Ipris_shell1_id_mean	0.039	0.649	61.8	76
		(0.534–0.752)
Ipris_shell1_id_min	0.045	0.635	81.8	40
		(0.520–0.740)
Ipris_shell2_gd_max	0.004	0.692	72.7	64
		(0.579–0.791)
Ipris_shell2_gd_min	0.010	0.679	81.8	56
		(0.565–0.779)

Abbreviation: AUC, area under the curve.

Figure 4 Receiver operating characteristic curves of the top three features for distinguishing between benign and malignant testicular lesions.