Improved Assessment of Middle Ear Recurrent Cholesteatomas Using a Fusion of Conventional CT and Non-EPI-DWI MRI

Article Figures & Data

Download figure
Open in new tab
Download powerpoint
FIGURE.
Example of rMEC leading to a carotid canal erosion. Recurrent middle ear cholesteatoma (axial b=1000 non-EPI MR imaging, B, arrow) probably located (rated Likert 4) in the hypotympanum, on uncoregistered dataset (A, B, C, E, arrows). The hypotympanum location is clear (rated Likert 5), considering that the fused dataset (b=1000/CT in axial plane, D, arrow) and the vertical carotid canal lysis (A and D, arrowheads) can be seen. The perioperative surgical findings (F) confirmed the hypotympanum rMEC location (arrows) and the carotid canal lysis (arrowhead). In that case, the presurgical image fusion allowed the surgeon to adapt his approach, thus lowering the surgical risks.

Table 1:

MR acquisition data

	T1WI	T2WI	SE DWI	3D-T1 VIBE
TR (ms)	610	4460	2000	9.19
TE (ms)	14	95	105	3.34
FOV (mm)	150 × 150	166 × 129	220 × 220	172 × 200
NEX	2	2	8	2
Slice thickness (mm)	2	2	3	0.7
Acquisition time (min:sec)	2:11	2	3:12	3:23
Matrix size	358 × 512	344 × 384	144 × 192	213 × 288
B-values			b=1000

Table 2:

rMEC surgical location findings and MR imaging correspondence

Anatomic Location	No.^{^a}	PREV	UD Exact	FD Exact	UD Reliab	FD Reliab
Epitympanic recess	18	83%	94%	100%	83%	97%
Posterior epitympanic recess	18	78%	100%	100%	92%	100%
Aditus of mastoid antrum	18	72%	72%	100%	72%	94%
Tegmen antri	18	67%	67%	100%	64%	97%
Anterior epitympanic recess	18	67%	67%	100%	83%	83%
Tegmen tympani	18	61%	83%	100%	81%	94%
Mastoidectomy cavity	18	56%	78%	100%	89%	94%
Facial nerve: tympanic segment	18	44%	67%	100%	64%	89%
Promontory of tympanic cavity	18	33%	94%	100%	86%	97%
Oval window/stapes	18	33%	83%	100%	83%	97%
Mesotympanum	18	28%	72%	100%	69%	100%
Vestibule	18	22%	78%	100%	97%	89%
Facial nerve: second genu	18	22%	83%	100%	81%	97%
Sinus tympani	18	17%	83%	94%	86%	97%
Eustachian tube	18	11%	89%	100%	86%	100%
Dura mater	18	11%	89%	100%	94%	94%
Round window	18	6%	83%	100%	78%	100%
Carotid canal	18	6%	94%	100%	94%	97%
Facial nerve: geniculum	18	6%	89%	100%	83%	97%
Cochlea	18	0%	100%	100%	100%	100%
Temporal lobe	18	0%	100%	100%	97%	100%
Facial nerve: mastoid segment	18	0%	100%	100%	94%	100%

Note:—PREV indicates surgical prevalence; Exact, correctly identified and correctly rejected findings (ie, true-positive + true-negative considering surgical findings as the criterion standard); Reliab, reliability score.
↵a Number of lesions.

Table 3:

Patient data

	Mean	Range	(%)
No. of patients	20
Mean age (yr)	41.2	(11.1–70.9)
Female/male (%)	35/65
Surgery
No. of previous interventions	2.9	(1–8)
Right/left (%)	55/45
Time to recurrence (mo)	70.6	(9–430)
MRI to surgery (mo)	2.1	(0–10)
MRI
True-positive in DWI	18		90%
False-negative in DWI	2		10%
Fusion quality
Good	9		50%
Perfect	9		50%

Table 5:

Segmentation reproducibility on volume measurements

	ICC	ICC-CI	Preci	Interpret	P Value^{^a}
Obs1 (vol 1–vol 3)	0.93	0.82–0.97	8%	Almost perfect	<.001
Obs2 (vol 2–vol 4)	0.86	0.67–0.95	14%	Almost perfect	<.001
Unfused (vol 1–vol 2)	0.91	0.78–0.97	10%	Almost perfect	<.001
Fused (vol 3–vol 4)	0.95	0.88–0.99	6%	Almost perfect	<.001

Note:—Preci indicates precision of ICC estimate; Interpret, interpretation of ICC; Obs, observer; vol, volume.
↵a Statistical significance using an ANOVA model for ICC.