Clinical forensic height measurements on injured people using a multi camera device for 3D documentation

The documentation of incidents is an important procedure in forensics [1‐4]. In forensic medicine, measurements of body height and injury positions are essential for the description and evaluation of an injury [5, 6]. It has been shown that injuries at different heights on the body of a pedestrian involved in a traffic collision allow investigators to make assumptions about the vehicle [1, 7]. In this article, we used the multicamera device Botscan© (botspot GmbH, Berlin, Germany) (Photobox) to measure and document injury positions on living people.

In forensic medicine, the measurements of injury shape and positions often depend on the description of the examiner, who can only perform measurements during and after the examination based on the documentation that was performed, such as body schemes and photographs. An extensive literature search revealed no standard procedure describing how to accurately describe and measure the position of injuries. However, several references stated that the positions of injuries should be documented [1, 4, 6‐8]. According to internal guidelines of the Zurich Institute of Forensic Medicine, in cases of traffic collisions, the heights of injuries are measured with a tape measure drawn out from the floor to the injury (Fig. 1).

For an accurate measurement, a tape measure or folding ruler should represent a straight, vertical line from the ground to the injury. This is problematic because holding the tape or ruler by hand often results in a roughly estimated vertical line. Furthermore, the tape often cannot be held with the necessary pressure, as it is difficult to affix the tape on the ground and simultaneously hold it at the necessary height. Third, the reading must be made orthogonally to the tape, which can be difficult for the person already holding it. It can also happen that the tape may also be slightly bent along the body to achieve the necessary stability, which does not result in a straight line from the ground but rather a pathway along the bent tape. In addition to executing the procedure, the measurement also has to be documented. In an optimal case, the measurement is not just written down but also documented with a photo showing the position of the injury along the tape and on the body. Additionally, the documentation should be carried out swiftly for the benefit of all involved parties.

This study used a mannequin equipped with injury stickers measured with tape and compared them to measurements performed with a 3D model created using the Photobox [23]. Furthermore, two real cases were included, and these measurements were performed as part of the case documentation. The measurements of the two injured persons taken during the medical examination were compared to the measurements taken using the 3D model.

First, the coordinates of the scale bar targets were used to check whether the scaling of the 3D models was correct. For this purpose, the distance between the targets of the scale bars was calculated and compared to the known distances of the scales. It was found that the maximum discrepancy was 15.6 mm and that the average discrepancy was 6.9 mm. For additional control, the distances between the plane and each of the four targets on the center platform were calculated, which should be close to zero. The largest distance was 3 mm when the left foot was forward (Table 1). On average, the distance was 1.2 mm (Table 2).

After it was established that the scale was correct, the measurements performed for the injuries were compared to one another. For the mannequin, the average discrepancy between the medical examination and the 3D model for 22 injuries with top and bottom measurements was −6 mm (±5 mm). The measurements performed with the 3D model were larger than the measurements performed during the medical examination (Table 3).

For the real cases, it was found that the measurements were shorter when the person was stepping forward than when the person was standing up straight (Tables 4 and 5).

The Bland-Altmann plot for the mannequin shows that both methods yielded comparable results (Fig. 5). For the injuries on the legs, the variance in the discrepancies was smaller than the variance for injuries on the torso; overall, the measurements performed with the 3D model were systematically longer than the measurements performed during the medical examination. For the real cases, the Bland-Altman plot showed that the measurements were similar but had a larger variance than the measurements on the mannequin (Fig. 6).

In this article, we presented and tested the Photobox for the 3D documentation of injury positions and compared the results to measurements performed with tape measures during a medical examination procedure.

It was shown that the Photobox data can be used to establish the location of injuries. The variance of 5 mm for the mannequin is in an acceptable range considering the many factors that could negatively influence the measurements of a stable object. The main problems are that during the medical examination, it can be difficult to establish and reach the correct foot that is planted on the floor in a perpendicular position and draw out the tape measure to the injury in a straight vertical line. Furthermore, taking a vertical reading on the tape measure while ensuring that the tape is affixed to both the floor and the injury is difficult. Additionally, the point of measurement may not be defined and might differ between physicians for the same injury. Without this definition, the measurement is dependent on the physician, and if the process of measurement is changed in the future, the measurements cannot be repeated. With Photobox data, however, measurements can be repeated for injuries even when the measurement system changes. While the Photobox appears to be an effective tool for injury documentation, the processing of the data requires technical knowledge and therefore skilled users to perform the subsequent measurements. For the mannequin, the injuries were clearly marked with dots at the upper and lower boundary. These dots were used by the physician and for measuring on the 3D model, allowing comparison of the measurements independent of having an undefined injury boundary.

The real case examples show that there are discrepancies between medical examiner measurement and measurements performed on the 3D models; however, the Bland-Altman plot indicates that the measurement discrepancies are well within a 95% agreement limit. One problem might be that the boundary of injuries is not always well defined, and medical examiners and persons performing 3D measurements might have different understandings of the injury boundary. Another error is human motion, as the persons had to move between the medical examination and 3D documentation, during which they altered their body positions and therefore the measurements. This also suggests that foot sole measurements, independent of the measurement method, can be different from those at the time of the actual event. Another finding with the real case examples was that the center platform moves depending on the foot position and stance, which influences the ground plane and therefore the subsequent measurements. This can be solved by documenting the persons in a defined shoulder width stance. The real case examples also posed another problem for the reassessment of nonvisible injuries, such as swelling. During the medical examination, the physician documented turgor at heights of 0.44–0.54 m, which was visible neither on the 3D model nor on the documentation photograph (Fig. 7).

The Bland-Altman plots show that for both mannequins and real cases, the measurements made with the Photobox are comparable to measurements made during the medical examination. This allows us to state that the Photobox can be used to perform injury documentation and localization. Due to the unavailability of a more accurate method, it is not possible to say whether Photobox or medical examination measurements are more accurate, but we can affirm that both methods yield similar results. However, previous results suggest that the Photobox yields more accurate measurements [23]. Another limitation is that the measurements were only performed by one person for each method. This reflects what would be done in real situations, in which the medical examination is only performed by one medical examiner, and no repeated measurements are performed. This was also done for the measurements on the 3D model to keep both methods comparable, even though with the 3D model, multiple measurements between different persons could have been performed.

In addition to the price, another limitation of the Photobox is that the device is stationary and cannot be moved; thus, it is only possible to document people who can stand up and move to the location of the Photobox. Additionally, physical examination, such as pushing on an injury to differentiate skin reddening vs hemorrhage, is not possible with either the 3D model or the 2D photograph. Despite these drawbacks, the 3D documentation of injured persons not only allows precise measurements of injury location and dimension but could also allow the use of 3D models for subsequent 3D reconstructions, e.g. matching of injury-causing object to injury [26‐29].

In particular, the ease of use, the speed and repeatability of measurements and the possibility of reassessing the dimension and localization of injuries provides an advantage for future investigations. Furthermore, photogrammetric images taken by a photographer depend on the experience of the photographer, while the Photobox acquires standardized photos under optimal lighting conditions. Additionally, the Photobox documents the whole body at once in 3D, while forensic images often only show the area of measurement and not the complete tape, thus potentially missing errors that might have occurred during the measurement procedure. However, it still needs to be analyzed whether the Photobox data can be used not only for measurement but also for forensic medical examination of the type and severity of injury.

The Photobox has been proven to be an effective tool for documenting the location of injuries compared to the current method of foot sole measurements. The use of the multicamera device Photobox has considerable potential for medical forensic examinations and should be further analyzed in future research.