2.2 MODEL O BAG-OF-WORDS (BoW)
2.2.1 Detección automática de puntos de interés: descriptores SIFT
Root et al (1971, 1977) proposed that the examination of the forefoot to rearfoot relationship can be used to examine the forefoot and midtarsal joint. Firstly, to identify if there is a structural deformity of the forefoot in the frontal plane, and second, to measure the range of motion available at the midtarsal joint. Root et al (1971, 1977) described how the midtarsal joint is the mechanism joining the rearfoot and forefoot, with the position and motion of the former affecting the latter.
To conduct the examination of the forefoot to rearfoot relationship as described by Root et al (1977), the patient is in a prone position, and the subtalar joint is held in a neutral position. Pressure is applied to the fifth metatarsal to pronate the midtarsal joint around both axes, locking the forefoot against the rearfoot. The angulation of the forefoot relative to the rearfoot is then measured by placing a measuring device or goniometer on the plantar aspect of the forefoot. Root et al (1971) advocated the measurement of the whole forefoot inclusive of the first to the fifth metatarsal heads to measure the angle of the forefoot. Although, Root et al (1971) suggested that if the first or fifth metatarsals are not in the same plantar plane as the second to fourth
83 metatarsals, most commonly because of a deformity of the first ray (e.g dorsiflexed or plantarflexed), then the plantar plane of the second to fourth metatarsals should be used.
Root et al (1971) stated that conducting the examination of the forefoot to rearfoot relationship can be difficult and practical experience is required. Some of the difficulties highlighted by Root et al (1971) and more recent literature (Garbalosa et al 1994, Evans et al 2003, Diamond et al 1989, McPoil et al 1988, Buchanan and Davis 2005) include: the difficulty of maintaining the subtalar joint in a neutral position whilst holding the fifth metatarsal, and using the goniometer, or forefoot measuring device to take the a measurement. Evans et al (2003) and Buchanan and Davis (2005) described how contraction of the tibialis anterior by the patient when the examination is conducted will create an inverted forefoot, representative of a structural deformity of the foot. They (Evans et al 2003 and Buchanan and Davis 2005) also emphasised how the variation between assessors in how much pressure is applied to the fifth metatarsal can also significantly affect the position of the forefoot. Other difficulties of this examination appear to be generic sources of error for the majority of the biomechanical examinations of the foot described by Root et al (1971, 1977). These include the low reliability of placing the subtalar joint into a neutral position, and the use of the bisection line drawn onto the posterior aspect of the calcaneus to represent the frontal plane position of the subtalar joint.
The intra- and inter-assessor reliability of the examination of the forefoot to rearfoot relationship
84 Garbalosa et al (1994) all report good to excellent intra- and inter-assessor reliability when using both a goniometer, and or forefoot measuring device. Garbalosa et al (1994) reported very good agreement between assessors using a goniometer (r = >0.894), or a forefoot measuring device (r = >0.929), the latter is what Root et al (1971) advocated. Evans et al (2003) reported ICC values of ICC = 0.823 for intra-, and ICC = 0.70 for inter-assessor reliability when examining the forefoot to rearfoot relationship in adults. In contrast, the ICC values calculated by Evans et al (2003) for children (ICC = 0.28), and adolescents (ICC =0.53) indicate poor to moderate reliability. In agreement with Lorimer et al (2007), this suggests that children’s feet are more difficult to examine, possibly due to the difficulty in maintaining the foot position when the measurement is being taken, and the small size of the feet. Although the reliability indices indicate good reliability, Evans et al (2003) highlighted howthe SEM values (SEM = 2.1 (adult)) are large considering the small mean result (2.01°). Diamond et al (2003) reported a similar result with ICC values of ICC = <0.93 for intra, and ICC = <0.77 for inter-assessor reliability. However, the mean angle of the forefoot to rearfoot relationship measured by Diamond et al (1989)was in a varus direction, but the large mean standard deviation values of SD =3° suggest there is large variation in the measurements between assessors.
The Root et al (1971, 1977) classification of the forefoot as varus or valgus
Root et al (1971, 1977) referring to Hlvac (1970) and Steindler (1929) described how a foot is classified with a forefoot varus if the forefoot is inverted, or a forefoot valgus if the forefoot is everted relative to the rearfoot. They proposed that to compensate for these deformities, the subtalar joint will abnormally pronate during
85 the gait cycle. Root et al (1977) stated that the close relationship between the subtalar and midtarsal joints allowed the subtalar joint to control the movement of the forefoot and restore a plantigrade contact of the foot with the floor.
In a foot classified with a forefoot valgus, the subtalar joint will remain in a pronated position during propulsion. In a foot classified with a forefoot varus, the subtalar joint will remain in a pronated position throughout the gait cycle. As the subtalar joint is in a pronated position, when it should be supinating, Root et al (1977) stated that the foot will be unable to transform into a rigid lever during midstance or propulsion. This will cause injury and deformity to the soft tissue and bony structures of the foot, in particular to the first metatarsophalangeal joint.
In agreement with Root et al (1977), Donatelli et al (1999) reported that in feet classified with a forefoot varus the calcaneus everted relative to the tibia during the contact, midstance and propulsion phases of the gait cycle. However, the cohort used by Donatelli et al (1999) included non-injured and injured professional baseball players and the frontal plane angle of the forefoot to rearfoot relationship was only 0.1° (p = >0.05) greater in the non-injured than the injured players. This suggests that this assessment provides little inference about the symptomology of the patient.
Furthermore, some (Garbalosa et al 1994, Buchanan and Davis 2005, McPoil et al 1988) have demonstrated that a large percentage of asymptomatic individuals can also be classified with a forefoot varus, or valgus. For example, McPoil et al (1988) classified from a cohort of 58 asymptomatic feet 44.8% with a forefoot valgus, and 8.62% with a forefoot varus. In contrast, Buchanan and Davis (2005) classified 92% of feet (n=43/51), and Garbalosa et al (1994) classified 86.6% (n=208/240) of feet with a forefoot varus. Both using large cohorts of participants. Although, McPoil et
86 al (1988) and Garbalosa et al (1994) followed the Root et al (1977) protocol and classified feet as a forefoot varus or valgus if the frontal plane angle of the forefoot was inverted or everted from 0°. Buchanan and Davis (2005) classified feet as a forefoot varus if the forefoot was inverted more than 8° from 0°. Feet classified with an inverted angle of between 1°-8° were classified as “neutral,” although Buchanan and Davis (2005) included them within the forefoot varus classification. Less than 1° inverted was classified as a forefoot valgus. There is a dearth of literature describing the kinematics of the foot in feet classified with a forefoot varus, or forefoot valgus. Results from the afore-mentioned investigations (Buchanan and Davis 2005, McPoil et al 1988, Garbalosa et al 1994, Donatelli et al 1999), strongly indicate that further investigation is required into understand the biomechanical function of feet classified with this type of structural deformity. It will then be possible to determine if the examination of the forefoot to rearfoot relationship is a useful predictor of injury.