Changes in electrical properties of bones as a diagnostic tool for measurement of fracture healing

Introduction In fractures, electrical properties are generated by the piezoelectric effect and cellular activity, initiate and augment healing. Monitoring these can result in the development a diagnostic tool for diagnosing delayed and early non-union of bones and may enable the clinician to change the line of treatment for decreasing the suffering time of the patient. This article summarizes 12 studies related to the electrical properties of bones for the monitoring of fracture healing. Materials and methods This experience has been used to develop a methodology comprising insulated fixators and measurement of electrical properties by an LCR (inductance, resistance and capacitance) meter at King George’s Medical University (KGMU). Inductance, conductance and impedance of the fractured and normal segments of fractured human tibia were monitored. Results Prospective data analysis was performed; this showed large variances. The patients were then stratified into two groups: (i) delayed union and (ii) normal union in a blinded manner. Analysis of data ensuring blinding was done separately. Conclusion Electrical properties are highly dependent on bone mineral density, temperature, structure and cross-sectional area of the bone. Skin and soft tissue are responsible for masking the electrical signals from bones measured in vivo. Therefore, at KGMU, insulated fixators were designed to prevent short circuit by rods and enable measurement from nothing else but the bone. Electrical properties of bones, can be used as a biomarker for monitoring fracture healing.


Introduction
Discovery of electrical potentials in bones dates back to 1957 as the piezoelectric effect 1,2 generated due to deforming forces on a hydroxyapatite crystal, which is also responsible for fracture healing [3][4][5] .In 1966, Friendenberg and Brighton observed that metaphysis of long bones is negatively charged, but the fracture site is more electronegative in nature 6,7 .This with other electrical properties generated due to cellular activity such as capacitance 8 , dielectric constant 9 , impedance 10,11 , inductance 12,13 and conductance 14 were experimented for developing a stimulation procedure for enhancing the healing rate 15 that could also be used as a diagnostic tool with time as the healing progresses 9,10,14 .Traditionally, bone union is considered as an end point 16 defined by the absence of abnormal mobility, absence of local tenderness and presence of transmitted movement and weight bearing 17 .However, in terms of patho-physiology, it is achieved by developing sufficient compressional and tensile strength across the fracture due to appropriate mixture of mineral and matrix deposition 18 , which is a continuous process, in the endosteal intermediate and parosteal zones of the callus 18 .Later, the parosteal and endosteal callus disappear and the intermediate callus becomes more mineralized to approximate normal cortical texture in the process of remodelling 18 .Thus, progression of fracture healing is not a yes-no end point decision and is subject to monitoring much before and after the end point is achieved 16 .Comparing the physical and electrical properties of the tissue in the fracture gap to the normal non-fractured adjacent bone fragments can form a useful parameter for monitoring fracture healing.Kooistra et al. summarized the various parameters required for a good measurement.As a 'gold standard' is not available, weaker hypothesis should be tested against the various available methods.The results should be validated in terms of construct validation based on established patient aspects and related clinician-assessed outcomes.The patient-related aspects can include psychotic distress, patient's judgement to the extent of weight bearing, physical functioning, body pain, mental and physical health (generic health-related quality of life measures), which can be judged by questionnaires like Short Form 12, SF-36, Health Utility Index and EuroQol dimensions, producing 243 classifiable different health states 17 .Similarly, clinical assessment outcomes can comprise assessment of abnormal mobility 18 , local tenderness 18 , weight bearing 18 , radiographic union scale of tibia (RUST) score 19 and radiographic assessment 20 for other bones except tibia, mortality, Disability Adjusted Life Years lost, reoperation and radiography 17 .Moreover, the resulting end point will not help the clinician to identify delayed and early non-unions while starting the antibiotic protocol, which can be made possible by measurements that are established against all patient-and clinician-related assessments.There is a need to cater to the loss of consensus among clinicians in the diagnosis of delayed, non-and mal-unions 21 .Therefore, the requirement comprises some digits stating at an early stage, the demarcation between normal and abnormal fracture healing.This paper summarizes all attempts that have been made in the direction comprising electrical parameters and suggests directions in which research should progress in this field.

Materials and methods
We systematically reviewed the literature, which summarized various experiments related to the electrical properties of bones.We performed a comprehensive search related to electrical properties of bones on Google, Google Scholar, MEDLINE, PUBMED, SCOPUS and all computerized electronic databases.All possible information was reviewed irrespective of the impact factor/indexing of the electronic databases.Articles that only reported data on electrical stimulation were excluded.All studies were independently assessed by two reviewers for inclusion eligibility and selected by two reviewers; disagreements were resolved by consensus.Each methodology is unique in its own way and provides information about the electrical atmosphere of bones and as the fracture heals.Their summarization and comparison among themselves and with the methodology developed at King George's Medical University (KGMU) as a part of a PhD thesis is an eye-opener for developing an instrument for measuring fracture healing.

Results
Twelve studies met our inclusion criteria; some assessed the electrical properties using in vitro methods, where as others used in vivo methods.In vitro experimentation gave the first line of evidence of usage of electrical properties as the fracture heals, followed by in vivo experimentation.Bones behave as a semiconductor due to positively charged collagen and negatively charged apatite, thereby creating a PN junction diode 22 .Water molecules bonded within the bone are responsible for this dielectric behaviour.The PN junction when subjected to infrared radiations produces photocurrent that increases in the presence of a magnetic field 22 .When osteoblasts are exposed to 72-Hz pulsed magnetic field for 10 min, they desensitize para-thyroid hormone on adenylate cyclase, resulting in increased collagen synthesis and decreased bone resorption by osteoclasts 23 .Measurement of magnetic properties is evident from the study by Kumatsu who measured the magnetic property in terms of bone inductance 13 .Cole and Baker observed inductance as negative capacitance.Pulsed electromagnetic fields have been used for bone regeneration and in bio-sensing devices for monitoring fracture healing 10 .For measuring electrical properties for diagnostic purposes, it is necessary to understand the electric field and current distribution of the whole bone with specific characterization of each part separately.Monitoring of difference in electrical properties of the marrow, cancellous and cortical parts is necessary for developing diagnostic procedures by impedance tomography.Conductance is a property of the bone marrow.Cancellous bone has lower resistance than cortical bone.They are highly frequency dependent, as resistivity of the cancellous bone decreases by 15% from 1 to 10 kHz when measured using an LCR meter 24 .Bone formation has also been monitored by tetracycline labelling and correlated with bioelectric potentials.A significant correlation exists in diaphysis and deviations that are present in metaphysis, showing that monitoring of healing by biopotentials at diaphysis is possible.Electrical potential generated in fractured bones (≤6 mV) can be masked by much higher potentials (>6 mV) generated in the damaged surrounding muscles.There is a need for a method to measure bio-potentials from the bone excluding the skin and soft tissue 25 .For this reason, insulated fixators were designed at KGMU, which permitted the measurement solely from the bone.Electrical properties were harvested from samples of the distal femur and proximal tibia and compared with mechanical properties, such as stress stain and Young's modulus, in the frequency of 50-5 MHz in vitro by an LCR meter.Permittivity and dissipation factors showed high correlation with Young's modulus, having negative correlations with conductance and positive with specific impedance 14 .Metaphysis is negatively charged in relation to diaphysis.When a bone is fractured, negative potential increased with the fractured area being most electronegative.When the fracture is healed, original charged distribution is restored.Denervation and blood flow do not affect bone potentials but depend upon cellular activity because cellular toxins cause profound electrical alterations 6 .
A technique for measurement of mutual inductance, small changes in impedance and resonant frequency was developed by Ko et al.-as the fracture heals, electrical impedance diminishes and electrical conductivity increases.The amplitude of oscillations in the detector peaks at the fracture site is highest when the fracture is new.As the fracture heals, there is a change in polarity, with the peak becoming a trough.These changes suggest that resonant frequency indicates progress in fracture healing 26 .
Impedance was calculated by output voltage and current measurements with the help of an electrical stimulator and oscilloscope in an in vivo experiment on distal radius fractures at regular intervals of 1 week.The mean Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY) Competing interests: none declared.Conflict of interests: none declared.
All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.For citation purposes: Gupta K, Gupta P, Singh GK, Kumar S, Singh RK, Srivastava RN.Changes in electrical properties of bones as a diagnostic tool for measurement of fracture healing.Hard Tissue.2013 Jan 21;2(1):3.
impedance ratio recorded postoperatively was 1.29 ± 0.3 at the second week, 1.36 ± 0.25 at the third week, 1.59 ± 0.37 at the fourth week and 1.65 ± 0.44 at the fifth week, which was significantly higher than that at week 1 in patients with normal healing.In patients with abnormal healing due to various complications such as loss of the second metacarpal pin and pin tract infection, impedance at the time of fracture union was lower than that at the first week.The methodology lacked insulated pins and measurement was taken with noise from the skin and soft tissue 10 .
Impedance imaging has been experimented and applied for measurement of bone healing and diagnosis of nonunions at an early stage 27 .Characterization of various bone tissues by multifrequency systems can also be beneficial 24 .A broadband signal for measuring electrical impedance containing maximum length sequence system and linear time variant systems can also be used to give additional spectral information necessary for diagnosing early non-unions 28 .However, further research is required to reduce the noise.
For knowing the exact time of removal of the fixator, the relationship of impedance values with callus maturation was studied in rabbits.Impedance was measured by an alternating current device at a frequency of 2 ± 0.4 Hz.Current was 30 ± 6 µA and loading resistance was 0-60 kΩ.Pins were not insulated; therefore, lot of noise from muscles would have been measured.Bone resistivity was calculated using bone impedance, length and mean cross-sectional area.A negative correlation between cross-sectional area and maximum bending stress was observed.Impedance significantly increased from week 1 to 6, with a decrease in the cross-sectional area.The pins used as a probe for the measurement were already used for fixation of the fracture.Impedance 11 was also measured in rabbits by Yoshida et al.Impedance with resistivity was monitored with an uninsulated fixator in a similar manner.They also suggested the monitoring of crosssectional area.However, they would also have easily monitored the electrical properties of the non-fractured segment of the fractured tibia.The cross-sectional area was monitored after excising the animal, which is impossible in case of human patients due to ethical reasons 29 .In vivo calculation of resistance by the direct current method as the fracture heals was performed on 12 patients with the help of current and voltage combination measurements using Ohm's law, and compared by fracture union assessed by radiographs.The fractures were fixed by a carbon ring fixator assumed to be non-conducting 30 ,but it was conducting due to the valence property of the carbon atom ring 31,32 .Blinding and RUST score were not used for radiographic assessment which has 50% accuracy 19 .The resistance of intact bone is 32.143 kΩ, which increases at the time of fracture to 115.38 kΩ 33 .Using alternating current for stabilizing ionic movement should be appropriate.On the basis of these observations, a methodology was developed at KGMU, Lucknow, India for monitoring fracture healing.

Methodology developed at KGMU
An external fixator containing rods and pins was re-designed.The sharp edge of the pins was filed and made blunt to prevent them from emerging through the second cortex when inserted.The pins and rods were coated by an insulating polymer, complying to USP Class VI biocompatibility standards for medical devices and implantation which did not permit conduction from skin and soft tissues.Coating was absent inside the bone permitting movement of current and measurement of the electrical properties of bone devoid of noise from skin and soft tissue as the healing progresses.Patients of compound fracture (Gustillo grade I and II) were treated with this electrically non-conductive fixator.
Electrical properties were measured every second week till 8 weeks at 100 Hz in 14 patients using an LCR-Q meter (Figure 1).Clinico-radiological parameters were recorded after the removal of the fixator at week 10 at every 2 weeks till the twentieth week.
The tibia was hypothetically divided into three segments: two controls and one fractured.The pins behaved as electrodes permitting measurement of the fractured segment as well as that of the controls, i.e. non-fractured segments of the fractured tibia.This method permitted the measurement of electrical properties in terms of conductance and impedance and electro-magnetic properties in terms of inductance.These parameters were divided by the length of the segment measured on non-digital radiographs.
Paired analysis for small sample (14 patients) comparing means of fractured segments separately with each proximal and distal segment was performed.Graphs were plotted of individual patients to compare the electrical properties of the fractured segment with the two controls at every 2 weeks till 8 weeks.Crude values are also reported.
Impedance and conductance showed high variance.Distal and proximal segments are not involved in fracture healing, yet they showed high variability in electrical properties as shown in Tables 1 and 2. The mean difference in impedance increased as the healing progressed, as shown by the P values (Table 2).At week 6, the difference decreased, and they sharply increased at week 8 denoting fracture union.

Conductance
Conductance can be a marker of fracture healing at week 8.There is no significant difference between the values of conductance at week 4 and 6, but at week 8, it is not statistically significant and tends to approach the cut-off point.A higher sample size is required to obtain a generalizable data.

Licensee OA Publishing London 2013. Creative Commons Attribution License (CC-BY)
Competing interests: none declared.Conflict of interests: none declared.

All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.
All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.For citation purposes: Gupta K, Gupta P, Singh GK, Kumar S, Singh RK, Srivastava RN.Changes in electrical properties of bones as a diagnostic tool for measurement of fracture healing.Hard Tissue.2013 Jan 21;2(1):3.
These two electrical parameters are dependent on the cross-sectional area and volume and bone mineral density (BMD) of that particular part of the bone.The proximal and distal segments are not similar in crosssectional area, volume and BMD as evident from the structure of the tibia and difference in the density of metaphysis and diaphysis.These facts suggest the need of monitoring and standardizing the electrical values based on volume and cross-sectional area.These readings are an eyeopener of the do's and don'ts for developing an instrument to monitor fracture healing.
Inductance was also measured.Inductance is independent of the cross-sectional area and volume and dependent only on the length of the bone fragment.It was standardized by dividing the inductance values with the length of the segment.
Inductance too showed high variance.However, when the graphs were plotted with inductance per unit length (Figure 2), the trend lines of the proximal and distal segments overlapped each other.There was a significant increase in inductance in the fractured segmentat week 4.
There was a significant difference in crude values of inductance of different patients.However, the graph of inductance per unit length versus time  showed significant similarity in trends in different patients (Figure 2).After data collection, the patients were followed up for 5 months.The patients were followed up at every 2 weeks and assessed by a blinded clinician for fracture healing on the basis of clinico-radiological parameters.They were classified into two groups, delayed union and normal union, by the blinded clinician.The data were analysed in a retrospective manner.Differences of mean and P-values were calculated.The impedance showed significant difference between the two groups at week 1, 4 and 6.Conductance showed a nearmarginal difference at week 6.This suggests that inductance and conductance can be used for the early diagnosis of delayed unions.The difference in the two groups was also assessed by the clinico-radiological parameters.In the normal union group, after removal of fixator at week 10, local tenderness was consistently absent, RUST score was higher at week 16 onwards (P = 0.01), absence of abnormal mobility was 58% higher and 100% at week 12 (P = 0.05), presence of weight bearing was higher from week 16 (OR = 15, P = 0.03), presence of transmitted movement was 2.4 times higher at week 10 (95% CI = 0.17-34.93,P = 0.52) and was 100% at week 14.The analysis was done for 14 of 17 patients, as two suffered injury again at the second and fourth week and one suffered pin loosening with infection.

Blinding
Blinding has played a key role in making the methodology rational in all aspects and is therefore separately discussed.Here, the PhD student taking the readings was unaware of the X-rays.Clinicians assessing the patients on clinico-radiological parameters were blinded from the electrical readings.A co-supervisor classified the patients into two groups.Statistical analysis was conducted by two separate unrelated statisticians for prospective and retrospective data.The data were separately provided by each collector for successful blinding to the respective statistician; the data was then compiled and analysed.

Discussion
The experiment designed at KGMU was able to caterto noise in the recording of bio-potentials.In the absence of noise, the data suggest

Figure 1 :
Figure 1: The set-up.The image on the left shows the X-ray of a patient with compound fracture.Both bones of the leg (Gustillograde I).Fracture has been fixed with insulated external fixator.A, B, C and D are insulated Schanz pins connected with insulated rods.Bone segment between pins A and B represents Control 1 (proximal segment), that between B and C represents Control 2 (fractured segment) and that between C and D represents Control 3 (distal segment).Note that the pins have blunt ends and do not protrude out of the cortex.The image on the right shows a clinical photograph during measurement of the electrical properties of the bone using an LCR-Q meter.

Figure 2 :
Figure 2: Graph showing inductance of the two patients.X axis: Time in weeks; Y axis: Inductance of each patient in µHenry.Control 1: Proximal non-fractured segment.Control 2: Distal non-fractured segment.Fracture: Fractured segment.

Table 2 Mean values of impedance and conductance with time of 14 patients Bone segments Time
Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY)Competing interests: none declared.Conflict of interests: none declared.All authors contributed to the conception, design, and preparation of the manuscript, as well as read and approved the final manuscript.All authors abide by the Association for Medical Ethics (AME) ethical rules of disclosure.For citation purposes: Gupta K, Gupta P, Singh GK, Kumar S, Singh RK, Srivastava RN.Changes in electrical properties of bones as a diagnostic tool for measurement of fracture healing.Hard Tissue.2013 Jan 21;2(1):3.

Table 3 Comparing delayed union with normal union (only P-values reported for comparison of electrical properties) Week Impedance Conductance Absence of abnormal mobility
Licensee OA Publishing London 2013.Creative Commons Attribution License (CC-BY)