Having undergone some physical therapy myself, I can confirm that even though keeping to a prescribed exercise and treatment regimen helps the injury heal faster, the daunting prospect of self-administered discomfort has led me to ignore the advice of my practitioner. A team at Microsoft's research wing has developed a working prototype of a system that may help to encourage physical injury sufferers to do their exercises by giving them a clearer understanding of what's going on. A therapist would use the device to project a series of graphics of underlying bone, muscle tissue, tendons or nerves directly onto the body of a patient to help explain the nature of the injury and prescribe effective treatment. The device can also take photos during a consultation, which can be subsequently reviewed or printed out as a memory aid for the patient.

It is estimated that up to half of patients undergoing physical therapy for chronic conditions fail to comply with the recommended therapies, and effective communication between patient and practitioner is seen as a major influence for compliance with prescribed exercise regimens. The team of Amy K. Karlson and Daniel Wigdor from Microsoft Research, and PhD student intern Tao Ni from Virginia Tech's Department of Computer Science, has created a system that could help to enhance such a therapist-patient information exchange.

The AnatOnMe projection-based handheld prototype is made up of two parts. The first consists of an Optoma PK102 pico projector, a Microsoft LifeCam digital webcam and a FireFly MV USB near-infrared camera. The second is a modified Logitech R400 laser pointer which has had its red laser diode replaced by an IR laser diode, and some control buttons added. Both parts are connected to a laptop for processing.

In the hands of a therapist

The therapist controls system functionality with the aid of the laser pointer, which is detected and registered by the system's near-infrared camera. During a consultation, the practitioner can use the AnatOnMe's webcam to snap photos of a patient's injured area and then review using the projector to throw the images onto a nearby wall. As the injury is discussed, the pointer can be used as a descriptive aid or to annotate the projected image. The original or enhanced images can then be added to a patient's medical records.

The researchers put together a series of annotated graphic collections representing six injury types using stock graphics, three upper body and three lower body injuries that often require physical therapy. Using this library, the therapist can project images onto a patient's body, a mannequin or a wall, to help the patient better understand an injury through 3D visualization of the problem and then to detail a recommended course of treatment.

Rather than creating a complicated automated system to line up the image with the area of injury, the prototype relies on the therapist to match the two by line of sight.

As the therapist gives exercise instruction, the camera could be used to photograph the patient performing the recommended exercises, and these photos could be compiled into an instruction sheet and printed off for the patient to take away. It is hoped that giving patients a virtual view inside an affected area will encourage them to keep up their exercises.

Testing the prototype

After screening volunteers from the local community for suitability, 18 participants were chosen to study the effectiveness of image projection on three types of surface during simulated therapy consultations. The volunteers were then asked for their impressions. Both Likert-scale and user ranking results showed that patients seemed to respond better to imagery projected onto body or model, with the wall-based testing proving the least engaging.

The researchers did, however, identify a potential key challenge for future development of on-body projection systems - a small number of volunteers had difficulty manipulating themselves into the position required for effective viewing of certain images. In a genuine injury situation, this could prove to be quite a stumbling block. Moving the projection to a model or using the practitioner's own body could offer simple, low-tech solutions to such issues, but it remains an interesting problem.

The system was also evaluated by two physical therapists who both agreed that such a device would be beneficial to treatment, and improve a patient's understanding of an injury and what's needed to treat it. They did note, however, that the current system kept both hands fairly occupied during a consultation and so prevented any touch and manipulation of the patient's injured area.

The researchers report being very encouraged by the overall positive reception to AnatOnMe and, although no future development plans have been officially announced, say that the next logical stage would be to validate their findings in the real world, with actual patients and genuine injuries. Other design choices for the device might also be investigated, together with possible adaptations for more general medical use.

View gallery - 10 images