Determination of ACSAmax
The participants were placed in the prone position on the examination table of a 3.0 T MR system (Magnetom Skyra, Siemens Healthcare, Erlangen, Germany). Then, the right foot and ankle were encased in the ankle coils (Foot/Ankle coil, Siemens Healthcare), and the ankle joint was positioned at 90° of plantarflexion (neutral position) using a Velcro strap to reduce motion artefacts. In accordance with the previous studies5,15, serial longitudinal relaxation time (T1)-weighted MR foot images of the range from the posterior part of the calcaneus to beyond the metatarsophalangeal (MTP) joint were acquired using a fast spin-echo sequence, with the following parameters: repetition time = 700 ms, echo time = 12 ms, average = 3, slice thickness = 3.5 mm, gap between slices = 0 mm, field of view = 125 × 125 mm, flip angle = 120°, and matrix = 512 × 512. To obtain MR images of the right lower leg, the participants were positioned on the examination table with their lower legs placed parallel to the main magnetic field. Serial lower leg images were acquired from the knee cleft just proximal to beyond the medial/lateral malleolus using the following sequential parameters: repetition time = 700 ms, echo time = 9.4 ms, average = 3, slice thickness = 5 mm, gap between slices = 5 mm, field of view = 360 × 360 mm, and flip angle = 120°, matrix = 512 × 512. The data acquisition time for each foot and lower leg scan was approximately 4 and 5 min, respectively.
Every MR image of the foot and lower leg was analyzed by one examiner (YK) using a specially designed image analysis software (SliceOmatic, version 5.0 Rev-3b, Tomovision Inc., Montreal, Canada). Non-contractile tissues, including the bone, tendon, fat, connective tissue, nerve tissue, and blood vessels, were carefully excluded wherever possible. From the foot images, the ACSAs of seven plantar intrinsic foot muscles, including the abductor digiti minimi (ABDM), abductor hallucis (ABH), ADDH-OH, ADDH transverse head (ADDH-TH), FDB, FHB, and quadratus plantae (QP), were manually outlined in every image from the most proximal to the distal image in which the muscle was visible. Of the plantar intrinsic foot muscles, other muscles that could not be visually separated from each other (i.e., lumbricals, flexor digiti minimi, and plantar interossei) were excluded from the analysis. The distance between the medial calcaneal tuberosity and the sesamoids bone of the first metatarsal of MR images was defined as the foot length (FL)5,15,16. The ACSAs of the plantar intrinsic foot muscles at 5% intervals along the FL were calculated by using spreadsheet software (Microsoft Excel, Microsoft Corp., Redmond, WA), and expressed relative to the FL (0% FL: medial calcaneal tuberosity, 100% FL: sesamoids bone of the first metatarsal). The maximal ACSA value for each individual plantar intrinsic foot muscle along the FL was defined as the ACSAmax. Furthermore, the ACSAs of all analyzed plantar intrinsic foot muscles, intrinsic great toe (FHB, ABH, ADDH-OH, and ADDH-TH), and lesser toes flexors (QP, FDB, and ABDM) were obtained for every 5% FL, and the maximal ACSA value of each functional muscle group was determined as the ACSAmax. Furthermore, the ACSAs of the extrinsic toe flexors (FHL and FDL) were analyzed from the lower leg images, and the maximal ACSA value was defined as the ACSAmax. In our previous study5, the good-to-excellent intra-rater repeatability was confirmed for the analysis of the ACSAmax of individual plantar intrinsic foot muscles and extrinsic toe flexors.
Measurement of TFS
A custom-made toe push dynamometer (T.K.K. 1268, Takei Scientific Instrument Co, Niigata, Japan), previously used to measure the TFS-All in previous studies12–14, was applied to determine each TFS. We modified this device to measure the TFS-great and TFS-Lesser separately (detailed below). The overall procedure for measuring each TFS was in accordance with previous studies13,14. First, the participants sat upright on the chair and their right foot was fixed on the ground of the device, with the hip and knee joints flexed at 90° and the ankle joint in a neutral position (90° plantarflexed) (Fig. 1a). The toe(s) intended for force production were then placed on a fixed and angle-adjustable sensor plate. As such, all toes were placed on the angle-adjustable sensor plate for TFS-All (Fig. 1b), only the great toe only for TFS-Great (Fig. 1c), and the lesser toes for TFS-Lesser (Fig. 1d). For the latter two measurements, the toe(s) not intended to produce force (lesser toes for TFS-Great and great toe for TFS-Lesser) were placed on another metallic plate mounted outside of the device to avoid touching the sensor plate (Fig. 1c and d). In all TFS measurements, the angle-adjustable sensor plate was placed at 45° of dorsiflexion at the MTP joint, in accordance with previous studies13,14. Previous reports have shown that the highest TFS values were determined at 45° of dorsiflexion at the MTP joint and 90° of dorsiflexion at the ankle joint angles in all cases when TFS was produced by all toes1, the great toe, and lesser toes17. To avoid extraneous movements, the ankle and forefoot were fixed on the ground of the device using Velcro straps. To measure each TFS, the participants were instructed to cross their arms in front of their chest and press down the sensor plate as strongly as possible with the toes defined above, without any extraneous movements. During the measurements of the TFS-great and TFS-Lesser, the participants were required to pay attention only to the toe(s) intended for force production and not to produce force with the non-intended toes. They were further instructed not to lift any toes not intended for force production from the metallic plate in all the tests. Two or three trials of submaximal force outputs for the familiarization procedures were conducted before the measurements. After familiarization trials with at least a 3-minute rest period, the participants performed the task with the maximal force for at least 3 seconds. The maximum effort trials were repeated twice, with a rest period of at least 1 minute. The largest value was used for further analyses. The intraclass correlation coefficients (ICC) of the two test trials were calculated to assess the inter-rater repeatability of each TFS measurement. The ICC (1, 2) values were 0.984 for TFS-All, 0.975 for TFS-Great, and 0.971 for TFS-Lesser.