Research Article: Psychophysical Evaluation of the Capability for Phantom Limb Movement in Forearm Amputees

Date Published: May 26, 2016

Publisher: Public Library of Science

Author(s): Noritaka Kawashima, Tomoki Mita, Benjamin Thompson.


A phantom limb is the sensation that an amputated limb is still attached to the body and is moving together with other body parts. Phantom limb phenomenon is often described on the basis of the patient’s subjective sense, for example as represented using a visual analog scale (VAS). The aim of this study was to propose a novel quantification method for behavioral aspect of phantom limb by psychophysics. Twelve unilateral forearm amputees were asked to perform phantom wrist motion with various motion frequencies (60, 80, 100, 120, 140, 160, 180, 200, 220, 240% of preferred speed). The attainment of phantom limb motion in each session was rated by the VAS ranging from 0 (hard) to 10 (easy). The relationship between the VAS and motion frequency was mathematically fitted by quadric function, and the value of shift and the degree of steepness were obtained as evaluation variables for the phantom limb movement. In order to test whether the proposed method can reasonably quantify the characteristics of phantom limb motion, we compared the variables among three different phantom limb movement conditions: (1) unilateral (phantom only), (2) bimanual, and (3) bimanual wrist movement with mirror reflection-induced visual feedback (MVF). While VAS rating showed a larger extent of inter- and intra-subject variability, the relationship of the VAS in response to motion frequency could be fitted by quadric curve, and the obtained parameters based on quadric function well characterize task-dependent changes in phantom limb movement. The present results suggest the potential usefulness of psychophysical evaluation as a validate assessment tool of phantom limb condition.

Partial Text

Most amputees still have a sensation and perception of the missing part even after limb amputation. This phenomenon is widely known as “phantom limb” [1,2]. While some amputees have vivid kinesthesia for their phantom limb, previous studies have described others as having an awareness of the missing limb as clenched and paralyzed in a specific position [3–6]. The literature indicates that there is a large inter-individual difference in the modality and extent of phantom limb awareness. There are many elements that affect the phantom limb condition, for example, missing part of body, type of prosthesis, pre-amputation condition, etc. Therefore, it is quite difficult to characterize the phantom limb phenomenon. A traditional and conventional way to characterize phantom limb is by describing the patient’s subjective feeling as a case report and presenting case series. Almost all outcome measures employed in behavioral and clinical studies regarding the extent of awareness and condition of a phantom limb were subjective or analyzed on a psychological scale, such as the visual analog scale (VAS).

Twelve male unilateral upper limb amputees participated in the present study (age: 59.0 ± 16.59 yrs). Nine subjects had undergone unilateral forearm amputation following a work-related traumatic injury, two subjects as a result of car accidents, and one due to post-cancer surgery. The stump length (distal to proximal end of the radius) ranged from 88 to 209 mm. Each subject was interviewed by one author (M.T.). The interviews documented medical history, present residual-limb (stump) sensations and pain, condition of the phantom limb, and phantom pain. The experimental protocol of this study was approved by the research ethics boards of the National Rehabilitation Center for Persons with Disabilities (NRCD). Written informed consent was obtained from all subjects before participation in the study.

As shown in Fig 2, the attainment of phantom limb motion as reported by VAS systematically changed with motion frequency, and showed task-dependent differences among the three experimental conditions. Although the preferred speed of the phantom limb motion was different among subjects, the VAS profile in accordance with motion frequency could be well fitted by the quadric (inverted U-shaped) function. Table 2 shows the coefficient determination (R2) of the quadric function in each condition obtained from all subjects. This mathematical fitting explains almost 60–70% of variance in the relationship between VAS and motion frequency (Uni: R2 = .71±0.91, Mirror-: R2 = .62±0.23, Mirror+: R2 = .71±0.16). Since Subjects D and G were able to accomplish phantom limb movement with the highest VAS at all frequencies under the Mirror+ condition, R2 value was not available for this condition for these two patients.

The purpose of the present study was to propose a novel quantification method for the behavioral aspect of phantom limb phenomenon using psychophysical evaluation. We asked forearm amputees to assess their attainment of phantom limb motion using the VAS. As shown in Fig 2, VASs changed systematically with motion frequency. Although the preferred speed of the phantom limb motion differed among subjects, VAS profiles in accordance with motion frequency were well mathematically fitted by an inverted U-shaped quadric function. Most subjects reported an optimal frequency which enabled them to make phantom limb motion easier, while if the phantom limb motion was slower or faster than the preferred speed, the motion became a little difficult. This comment indicates that the quadric function shows good agreement with subjective feeling.