Uses of Phantom Sensations Induced by Global and Local Modifications of the Prosthetic Socket as Somatosensory Feedback During Walking in Lower Limb Amputees (PhantomWalk)

People without fantom sensation during gait

Participants with unilateral amputations and a control group will undergo a comprehensive walking analysis. In a lab setting (on a treadmill or floor), their gait will be recorded using inertial sensors, EMG, force plates, and motion capture, while an EEG cap and a NASA-TLX questionnaire will assess cognitive load. Capable participants will also perform a walking test on a real-world outdoor course with varied terrain. The data will be analyzed to compare walking patterns between groups (amputees vs. controls, and amputees with vs. without phantom limbs). Finally, participants will describe any interactions they perceived between their phantom sensations and walking during the tests.

Participants reporting referred sensations in gait analysis phase will undergo further testing. The study will investigate whether prosthesis pressure or a specific socket interaction causes these sensations. A 3x3 cm grid will be drawn on the residual limb to systematically map referred sensations using various stimuli (brush, pressure tip, electrical stimulation). Participants will rate the intensity and describe the location of any resulting phantom sensations. This mapping will be repeated after 6 months to assess the stability of these sensations over time, with the ultimate goal of potentially using them to provide sensory feedback for prosthesis users. The session will last approximately 2 hours.

A semi-structured interview will be conducted to ask participants to describe their phantom limb sensations (type, location, intensity) under three conditions: without their prosthesis, while standing with it, and while walking with it. This aims to identify factors influencing these sensations. Additionally, a brief physical examination of the residual limb will be performed to see if touch in specific areas triggers any referred sensations.

People with phantom sensations who are caregivers

Participants with unilateral amputations and a control group will undergo a comprehensive walking analysis. In a lab setting (on a treadmill or floor), their gait will be recorded using inertial sensors, EMG, force plates, and motion capture, while an EEG cap and a NASA-TLX questionnaire will assess cognitive load. Capable participants will also perform a walking test on a real-world outdoor course with varied terrain. The data will be analyzed to compare walking patterns between groups (amputees vs. controls, and amputees with vs. without phantom limbs). Finally, participants will describe any interactions they perceived between their phantom sensations and walking during the tests.

Participants reporting referred sensations in gait analysis phase will undergo further testing. The study will investigate whether prosthesis pressure or a specific socket interaction causes these sensations. A 3x3 cm grid will be drawn on the residual limb to systematically map referred sensations using various stimuli (brush, pressure tip, electrical stimulation). Participants will rate the intensity and describe the location of any resulting phantom sensations. This mapping will be repeated after 6 months to assess the stability of these sensations over time, with the ultimate goal of potentially using them to provide sensory feedback for prosthesis users. The session will last approximately 2 hours.

A semi-structured interview will be conducted to ask participants to describe their phantom limb sensations (type, location, intensity) under three conditions: without their prosthesis, while standing with it, and while walking with it. This aims to identify factors influencing these sensations. Additionally, a brief physical examination of the residual limb will be performed to see if touch in specific areas triggers any referred sensations.

In 10 unilateral transfemoral amputees with stable induced phantom sensations, this phase evaluates how a personalized socket modification affects phantom sensations and walking. With the prosthetist, a localized thickness (e.g., silicone insert) is added inside the usual socket at the stable stimulation sites identified on the residual-limb map, after measuring socket pressure with a thin sensor array. Each participant is tested in two conditions in a fixed order: usual socket first, then modified socket. The order is not randomized because induced sensations may persist after the modification is removed. For each condition, gait is analyzed in the laboratory and on the ecological course (4 sessions; conditions no more than 4 weeks apart). Participants describe their sensations, rate cognitive load, effort and socket comfort, and indicate which condition is most helpful for walking.

Asymptomatic people

Participants with unilateral amputations and a control group will undergo a comprehensive walking analysis. In a lab setting (on a treadmill or floor), their gait will be recorded using inertial sensors, EMG, force plates, and motion capture, while an EEG cap and a NASA-TLX questionnaire will assess cognitive load. Capable participants will also perform a walking test on a real-world outdoor course with varied terrain. The data will be analyzed to compare walking patterns between groups (amputees vs. controls, and amputees with vs. without phantom limbs). Finally, participants will describe any interactions they perceived between their phantom sensations and walking during the tests.

People with bothersome phantom sensations

Participants with unilateral amputations and a control group will undergo a comprehensive walking analysis. In a lab setting (on a treadmill or floor), their gait will be recorded using inertial sensors, EMG, force plates, and motion capture, while an EEG cap and a NASA-TLX questionnaire will assess cognitive load. Capable participants will also perform a walking test on a real-world outdoor course with varied terrain. The data will be analyzed to compare walking patterns between groups (amputees vs. controls, and amputees with vs. without phantom limbs). Finally, participants will describe any interactions they perceived between their phantom sensations and walking during the tests.

Participants reporting referred sensations in gait analysis phase will undergo further testing. The study will investigate whether prosthesis pressure or a specific socket interaction causes these sensations. A 3x3 cm grid will be drawn on the residual limb to systematically map referred sensations using various stimuli (brush, pressure tip, electrical stimulation). Participants will rate the intensity and describe the location of any resulting phantom sensations. This mapping will be repeated after 6 months to assess the stability of these sensations over time, with the ultimate goal of potentially using them to provide sensory feedback for prosthesis users. The session will last approximately 2 hours.

A semi-structured interview will be conducted to ask participants to describe their phantom limb sensations (type, location, intensity) under three conditions: without their prosthesis, while standing with it, and while walking with it. This aims to identify factors influencing these sensations. Additionally, a brief physical examination of the residual limb will be performed to see if touch in specific areas triggers any referred sensations.

In 10 unilateral transfemoral amputees with stable induced phantom sensations, this phase evaluates how a personalized socket modification affects phantom sensations and walking. With the prosthetist, a localized thickness (e.g., silicone insert) is added inside the usual socket at the stable stimulation sites identified on the residual-limb map, after measuring socket pressure with a thin sensor array. Each participant is tested in two conditions in a fixed order: usual socket first, then modified socket. The order is not randomized because induced sensations may persist after the modification is removed. For each condition, gait is analyzed in the laboratory and on the ecological course (4 sessions; conditions no more than 4 weeks apart). Participants describe their sensations, rate cognitive load, effort and socket comfort, and indicate which condition is most helpful for walking.

People with neutral phantom sensations

Participants with unilateral amputations and a control group will undergo a comprehensive walking analysis. In a lab setting (on a treadmill or floor), their gait will be recorded using inertial sensors, EMG, force plates, and motion capture, while an EEG cap and a NASA-TLX questionnaire will assess cognitive load. Capable participants will also perform a walking test on a real-world outdoor course with varied terrain. The data will be analyzed to compare walking patterns between groups (amputees vs. controls, and amputees with vs. without phantom limbs). Finally, participants will describe any interactions they perceived between their phantom sensations and walking during the tests.

Participants reporting referred sensations in gait analysis phase will undergo further testing. The study will investigate whether prosthesis pressure or a specific socket interaction causes these sensations. A 3x3 cm grid will be drawn on the residual limb to systematically map referred sensations using various stimuli (brush, pressure tip, electrical stimulation). Participants will rate the intensity and describe the location of any resulting phantom sensations. This mapping will be repeated after 6 months to assess the stability of these sensations over time, with the ultimate goal of potentially using them to provide sensory feedback for prosthesis users. The session will last approximately 2 hours.

A semi-structured interview will be conducted to ask participants to describe their phantom limb sensations (type, location, intensity) under three conditions: without their prosthesis, while standing with it, and while walking with it. This aims to identify factors influencing these sensations. Additionally, a brief physical examination of the residual limb will be performed to see if touch in specific areas triggers any referred sensations.

In 10 unilateral transfemoral amputees with stable induced phantom sensations, this phase evaluates how a personalized socket modification affects phantom sensations and walking. With the prosthetist, a localized thickness (e.g., silicone insert) is added inside the usual socket at the stable stimulation sites identified on the residual-limb map, after measuring socket pressure with a thin sensor array. Each participant is tested in two conditions in a fixed order: usual socket first, then modified socket. The order is not randomized because induced sensations may persist after the modification is removed. For each condition, gait is analyzed in the laboratory and on the ecological course (4 sessions; conditions no more than 4 weeks apart). Participants describe their sensations, rate cognitive load, effort and socket comfort, and indicate which condition is most helpful for walking.

Proportion of amputee participants able to walk who report non-painful phantom sensations during walking, collected during a standardized semi-structured interview. Reported as a proportion with its exact (Clopper-Pearson) 95% confidence interval

Double support time, defined as the sum of the initial and terminal double support periods of the prosthetic-limb gait cycle, expressed as a percentage of gait cycle duration (%). Measured by 3D motion analysis and force plates during level overground walking at self-selected speed in the laboratory (spontaneous condition).

Difference in double support time (% of gait cycle, same definition and walking condition as Phase 2) between walking with the usual socket and walking with the socket modified by a focal thickness addition designed to induce phantom sensations, in transfemoral amputees. Conditions assessed in a fixed order (usual socket first). Reported as mean difference with 95% confidence interval.

The number of steps per minute (steps/min) when walking at a comfortable speed. Measured using inertial measurement units or the Vicon system.

Electroencephalographic (EEG) activity will be recorded via a headset. Cognitive load will be quantified by analyzing the power ratio in specific frequency bands at the frontal electrodes.

The average walking speed (in m/s) measured using inertial measurement units and pressure soles (Loadsol) during a predefined outdoor route including various difficulties (stairs, slopes, etc.).

Participants' verbal descriptions of the nature of the interaction (facilitating, hindering, neutral), the evolution and characteristics of phantom sensations during walking will be collected via a structured interview at the end of the session. The data will be analyzed thematically.

For the subgroup of participants reporting interaction between sensations and focal stimulation after various stimulations using a rubber tip (pressure), a brush, and a needle. The change in sensation intensity (rated on a scale of 0 to 5). The result will be categorized as "increase," "decrease," or "no change."

Perceived mental workload will be assessed using the NASA-TLX questionnaire immediately after each walking session. The total score will be compared between the four socket conditions.

The overall comfort perceived for each condition will be assessed at the end of each session using a Numerical Rating Scale (NRS) ranging from 0 (very uncomfortable) to 10 (very comfortable). The average score will be compared between the four conditions.

The cognitive load perceived by the participant will be assessed using the NASA Task Load Index (NASA-TLX) self-report questionnaire. The total score, ranging from 0 to 100, will be calculated. A higher score indicates a greater perceived cognitive load.

For each condition, a structured description of the sensations reported (type, location, intensity, perceived usefulness for walking) will be collected through an interview. The data will be analyzed thematically to compare the sensory experience.

For the subgroup of participants with an amputation of less than 6 months at inclusion, the intensity of phantom sensations will be re-assessed using the 3-point ordinal scale.

The participant's average walking speed, measured in meters per second (m/s). The measurement will be collected using inertial measurement units (IMUs) and/or a Vicon motion capture system while the participant walks on hard ground or on a treadmill at their self-selected speed.

Gait symmetry will be calculated based on the stride length of the affected (or corresponding) limb and the healthy limb. The symmetry index is a unitless ratio (or percentage) that quantifies the difference between the two limbs. A value of 1 (or 100%) indicates perfect symmetry. Measured using inertial measurement units or the Vicon system.

For each type of stimulation, the number of 3×3 cm areas of the residual limb eliciting phantom sensations is counted and the total surface area (cm²) is calculated (number × 9). For each reactive area, the participant rates the intensity of the induced sensation on a three-level ordinal scale (weak, medium, strong).

At the end of Phase 4, participants indicate which condition (usual socket or modified socket) they perceive as most facilitating for walking.

Walking speed (m/s), cadence (steps/min) and step symmetry index measured with inertial sensors and pressure insoles during the predefined indoor/outdoor course, compared between the usual socket and the modified socket.

Composite assessment of phantom limb sensations via a semi-structured interview: type of sensation, location, intensity (weak, medium, strong) and perceived influence on walking (helpful, neutral, disturbing).

Step length (m) of the prosthetic and intact limbs during level walking at self-selected speed.

Stance and swing phase durations of each limb, expressed as % of gait cycle.

Composite score computed from spatiotemporal parameters, ranging from 30 (highly pathological gait) to 100 (healthy gait).

Composite gait variability score computed from step-to-step spatiotemporal parameters; values between 95 and 105 indicate healthy variability.

Perceived cognitive workload during walking, computed as the mean of the six NASA-TLX dimensions, each rated 0-10 (higher = higher workload).

Borg CR10 score (0-10; higher = greater exertion), collected after each walking condition.

Visual analog scale from -5 (very disturbing) to +5 (very helpful), completed by participants with phantom sensations after each walking condition; analyzed as a continuous variable

The maximum force exerted on the ground with each step, measured in Newtons (N) or as a percentage of body weight (%BW). The measurement will be taken using force platforms (AMTI), pressure insoles (Loadsol), or an instrumented mat (Zeno).