Study sample description
A retrospective review has been launched to 18 medical records of consecutive patients with TON in the Department of Otolaryngology–Head and Neck Surgery, Xiangya Hospital, from January 2017 to July 2021. Indirect TON was diagnosed by a history of blunt head or facial trauma combined with decreased visual acuity, color vision, and a relative afferent pupillary defect with a normal fundus during the early period of post-trauma . We excluded cases with direct trauma to optic nerve identified by high-resolution CT scan of the orbital and optic canal. The patients with indirect TON were intravenously administered methylprednisolone (30 mg/kg per day) every day for three consecutive days as previous studies described . Since there was no consensus on grading criteria in evaluating the severity of indirect TON, TON with best-corrected visual acuity equal to or more than 1.85 logMAR after EOND surgery was considered as severe TON, and those patients were recruited in this study. All of the subjects received SWD therapy and perceptual training after EOND surgery.
Shortwave diathermy therapy
DL-C Ⅱ (Dajia®, Shantou, China) SWD device was applied in this study, and therapeutic apparatus was delivered at a frequency of approximately 27.12 ± 0.6% MHz in continuous modes with a power of 50 W ± 20%. In our study, two sets of circular capacitive electrodes were used with a diameter of 80 mm for adults and 50 mm for children. The fracture areas were located using high-resolution computed tomography (Fig. 1). One electrode was applied in the temporal area of the ipsilateral side of the injured nerve identified by a three-dimensional alignment procedure through high-resolution computed tomography scan. This area was not only a shorter distance from the skin to the injured nerve but was also as far away from the crystalline lens as possible. To decrease the negative effects of SWD on the nerve systems , another circular electrode of SWD was applied in the frontal area of the ipsilateral side of the injured nerve (Fig. 2). In the acute stage of the TON (within 10 days after injury), the athermal mode of the SWD was applied to the marked areas for a 10-minute daily session. Ten days after injury, the microthermal mode was used for a 15-minute daily session. All subjects in the rehabilitation group received SWD therapy 5 days per week for 4 weeks.
Perceptive learning sessions were composed of a series of training procedures. First, sensory stimulations including touch, stroking, tapping, and pressing on the local area of the ipsilateral eye as well as the surface location of the injured optic nerve were offered. Second, light stimulation was applied at different intensities of bright based on evaluation of the pupillary light reflex of the subject, and avoiding glare and longtime light stimulation on the eye when training (Fig. 3. A). Additionally, the subjects were instructed to enter the bright room from the dark room when they were wearing an eye patch in the intact eye. Third, differentiation of the shapes and objects was performed in dynamic or static states with different colors, different sizes, and different shapes  (Fig. 3. B-C). Fourth, different written words with different sizes and different distances were discriminated in the different directions of the eye (Fig. 3. D). Fifth, color vision was trained using real objects or pictures (Fig. 3. E). Finally, the visual field training was scattered in the above training methods. Each training media including the light, objects, words, and color was input from the anterior, superior, inferior, nasal, and temporal sides of the eye. In addition, many training techniques were used in this study. First, visual imagery (VI) training was used when the patients had a complete or severe visual loss. The intervention of the VI training was as follows: the subjects used the intact eye to observe the object for 5 seconds, then they closed their eyes and imagined viewing the object using the injured eye for 10 seconds for 10 repetitions. Additionally, proprioceptive training, including going over the barriers, going up and downstairs, walking the slope, and training the balance, was executed. Finally, constraint-induced perceptive therapy was performed when the visual lesion partially recovered. The subjects wore an eye patch on the intact eye for 30 minutes in one section, 2 to 4 hours a day, 7 days per week (Fig. 3. F). All the subjects received perceptive learning therapy 5 days per week for 10 weeks except constraint-induced perceptive therapy.
All subjects underwent a thorough evaluation of visual function by an ophthalmologist preoperatively, postoperatively, and post-treatment. The ophthalmologic examinations consisted of visual acuity with optimal correction lenses for both eyes, color vision, pupillary light reflex, relative afferent pupillary defects, visual field examinations, visual evoked potential (VEP), and fundoscopy.
Best-corrected visual acuity was measured by a standardized Snellen visual chart. Visual results were converted into logMAR units for the convenience of statistical analysis. Hand motion, light perception, and no light perception were converted to 2.3, 2.5, and 3 logMAR units, respectively . Color vision was evaluated using the Ishihara color vision test 24 plate . The exclusion criterion of the color vision test was congenital color vision deficiency. The number of correct answers in a set of 24 plates was recorded as the color vision score. Relative afferent pupillary defects was evaluated by using a swinging flashlight with a grade of one to five. Pattern visual evoked potential testing was recorded using an MEB-9404C (Nihon Kohden Corp, Tokyo, Japan). The latency and amplitude of the P100 wave in the VEP testing were collected for analysis.
Additionally, a 32-channel head coil on a 3.0T MRI system (Philips, Ltd, Best, the Netherlands) was used for the acquisition of imaging data. T1- and T2-weighted, fat-suppressed images were obtained axially through the orbit and some parts of the brain, including the intracranial portion of the optic nerve, postoperatively and 10 weeks after rehabilitation. Optic nerve diffusion tensor imaging images were obtained using single-shot echoplanar imaging sequences. The images were acquired from the optic papilla to the orbital apex of the optic nerve with 40 contiguous slices. The following imaging parameters were used: acquisition matrix = 80 · 78, reconstructed to matrix = 128 · 128 matrix, field of view = 200 ·200 mm2, TR = 2214 milliseconds, TE = 82 milliseconds, parallel imaging reduction factor (SENSE factor) = 2, EPI factor = 39 and b = 800 s/mm2, NEX = 2, slice gap = 0, and a slice thickness of 3 mm. The region of interest (ROI) was manually placed over the optic nerve at the anterior, middle, and posterior segments on the non–diffusion-weighted (b0) image. The fraction anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity of optic nerves were measured.
All analyses were performed using SPSS version 22.0 software (SPSS, Chicago, Illinois, USA). After the data passed normality tests and homogeneity of variance tests, the mean, standard deviation (SD), and range values were calculated. We conducted the One-way analysis of variance (ANOVA) to compare the difference before and after treatment intervention for visual acuity and color vision, and the Student's t-test for VEP and diffusion tensor imaging parameters.
Pearson correlation was used to analyze the relationships between age, visual functions, diffusion tensor imaging, and VEP variables. A P-value <0.05 was considered statistically significant.